I love being a Data Scientist and have never had a better job. To me the job of a Data Scientist is to translate a business problem into a math problem which can be solved with the available data. Solving it is the easy part. To do this, we are part business analyst, part financial analyst, part data engineer, part software developer, part machine learning engineer, part statistician, part visualization engineer and part product manager. The ambiguity is off the chart and the need to learn is constant. When I hire, I look for determination, intellectual curiosity, desire to constantly learn as well as the ability to deal with ambiguity and failure. The hard skills I can teach.
Hello Rick, it's an interesting job based on your explanation. I work as a logistic business analyst and I've been trying to switch to data science and really need a good mentor to set my mindset in place. Could we connect? I'm really looking forward to hear your feedback 🙏🏻
I love this comment. As a data scientist, curiosity, enthusiasm is the key. For me 2 years now freelancing in the field and I know how it feels. Another important part, you must be a team player and in most cases have a good managerial skill in order to advance in the career. I wish I get the opportunity to work with you someday.
I'm learning Data Science and whenever I read job descriptions for entry levels or even for Interns they want you to be jack of all trades and on top of that require you to have 2,3 years of experience. Plus they include big data streamlines and ask you to be an expert on it too along with all sorts of cloud services. I mean it feels like they're demanding an Intermediate Data Scientist and terming it as "Data Analysis Intern" or "Junior Data Analyst" role. This frustrates me a lot.
In my personal experience the harsh reality also included: 1. Studying all the time, to the point that you personal life gets affected. 8 hours of work and 4 hours a day of just keep up with the new lake, cloud , software, technique, library , bi tool etc. 2. The feeling of "You don't know enough", "You are not enough". You briefly mention this and I am glad that I am not the only feeling the same way. I know python enough to do my job, but I am not a python expert who can write functions from scratch, or change between languages say from python to java or R. I use the existing libraries and move on to the next project.
You don’t need to keep up with the new stuff. It’s not like companies have unlimited money to switch cloud when a product releases. Or change a whole working thing because this bi tool is cool. Tf you work
Build foundation. If you understand the basic concepts of object oriented programming, basic computation, os (unix based kernels), then keeping up with trends isn't all that difficult. Technologies, libraries, frameworks, etc. are an abstraction of the fundamentals. You'll never "know" enough. It's not about knowing. It's about collaboration, researching the issue, and gathering the best tools to solve the problem.
Its really difficult to work as a Data scientist because every company have different expections from Data Scientist. And now a days, a data scientist should know all the skills like computer vision, deep learning, Operational research, Machine learning, SQL and many more. There must a discrete line between data scientist work roles. And also as you have mentioned interview prep, it is really difficult because we need to prep from scratch. - Struggle of a DataScientist
But many online institutes provide training saying anyone can be data scientist and get placement too.. with good CTC of minimum 7 L is that true, Do any one transist this career in Data scientist ? Is good option ?
I agree with the part about the necessity of being versatile. Of course, looking for a job, one can just ignore the job offers containing techs one does not know. But in my view, there are so many of them that after that 'offer elimination' process, one's final list could turn out to be pretty short. Right now, I work as a Mid Data Scientist, and although I feel confident in what I am doing on a daily basis, I have (and constantly update) my 'to learn' list.
I don't know whether to laugh or cry when you say you are expected to know Operational research as well as the other stuff. Operational Research is huge (not just optimisation). It certainly requires data analysis and statistics skills - and SQL and communications and modelling and data cleaning..... I found that people are in awe of the trivial stuff that they can almost understand (e.g. spreadsheets with actual formulae!) and the challenging stuff is so far over their heads they don't even realise it is needed or even that it is there. I can relate to Sundas Khalid when I think of managers I had who really did not appreciate what the discipline offers. I had to do the stuff they asked for while trying to slip in the stuff that I knew would be useful. Now I can see that the same applies to related topics such as Data Science.
Ok I thought I was crazy feeling like I had to learn financial risk modeling from scratch before my interview and do a little quick project to show my ability lol This being for a Quant risk analyst role at a bank
People usually leave managers, not jobs. Your experience there is not unusual. I also work in IT, and have resisted moving up to management. I'd rather be involved with the work, rather than stepping back to manage the workers. However, this has hurt my career in some work environments (Notably business, less research) where not wanting to move up into management is seen as a lack of motivation. So, for me, I've taken hits in compensation in order to support my own job satisfaction. I'm now old enough that retirement is on the horizon. I look at the size of my nest egg and contrast it with the ulcers that would have been required to make it larger, and I don't have many regrets.
So few people get this. It's not just about getting ulcers, but avoiding getting promoted to a level of incompetence. Not to offend anyone, but to be a manager is a completely different skill set that even many of those pursuing it don't fully grasp. Not to mention how even great managers aren't able to act beyond what they're allowed to do by, let's say, poor company policy. This is why there are so many inadequate managers present in so many workplaces across every possible industry imaginable. A good, great or awesome worker doesn't necessarily make a good manager. Also, being a manager isn't always more valuable than a skillful worker. (In any field) Thus, it's a shame there are examples such as yours where a person takes a hit in compensation or any other benefit for what is a silly reason.
Yes there’s ambiguity in role definitions. In my previous company, I worked as Data scientist. However I hardly did any data science work. It was more like a developer work where I was assigned to test and deploy the ML models in production. I could learn a lot about aws cloud architecture and services. Trouble shoot real time issues. But did little of ML. Then I changed my job and joined as Data analyst. Here I do little of analysis, SQL etc and more of ML like developing models and fine tuning them. Cleaning the data, feature engineering and feature selection. I would caution the people aspiring to enter the field because many a times companies themselves are not sure where and how they are going to use data science, what are the prerequisites for a successful data science project. They lack quality data, if they do have data- accessing and compiling the data is another issue. After aligning with multiple departments as single man army, you are expected to give some magical results. When results are satisfactory, you will be asked to deploy them. Here again the streamlined organisational requirements are missing. Somehow you drag the project. Later you are expected to monitor and troubleshoot issues in production. And then after all of this single man army efforts, management doesn’t see any value in it. Just to keep up with trend companies end up hiring a data scientist and expect that person to assume multiple roles and give them an end to end solution. They are just giving analytics a try without investing much in it.
Thank you for sharing! I have a really similar story. Worked as a CV/ML for a 3 years and constantly tried to be up to date with last tech stack and ml frameworks but only few time had real chance to use them. Now I'm in process to switching to frontend.
@@dariashtanakova1200 front end developer will be completely different field right ... You will need more coding html, java etc ... What are other options you can opt for?
I'm also a data analyst, after completing college degree applyed several job, after got rejected from more then 50 companies. Got i job in one company with heavy workload plus nightshift too. Then i decided to leave it. Then decided to change field. And now i started e-commerce online business and put my all computer skills in business. Most of my work done by AI. And i really happy with my decision
Great insight! I got into data science right after college and really struggled to do so. The fact that the data science job role is defined so differently at different companies made interviews really hard to prep for. It felt like you need to be a jack of all trades and be well versed in data science, data engineering and data analyst principles and it was just a lot to prep for. Plus a lot of roles require graduate degrees. After I got a role, I left after a year and transitioned into SWE as like you said the expectations didn’t match the reality plus I was also at a toxic workplace. I realized that I want to be in a role that is more well-defined and somewhat uniform across companies so if i need to look for a job later, I’m not being restricted to only a subset of opportunities bc of the different definitions and requirements.
Hey! I am actually trying really hard to get into data science but there are a few complications that i am facing and I don’t have a degree in this field, i spoke with a few counsellors as well but they couldn’t provide me with the answers that i was looking for, in the end I believe someone who is already in this field can help me out with my queries, so i was hoping if you could like help me a little on a few concepts?
Yeah! Totally agree. As a bioinformatician, I also felt some things this video reported. I think the "data scientist" role is even newer and less well-defined. So, managers and colleagues don't know what it is, which leads to great mismatches in terms of expectations... As far as industries evolve into more data-centric approaches, I guess people will be better valued...
The biggest problem I’ve had so far is working with my manager. We both have very different perspectives on various problems but he simply doesn’t even want to hear my perspective even when my perspective is usually right. We are both data scientists but he graduated as a data scientist back in 2003. So he’s an old head and I recently graduated. It’s very tough to get along and having your voice heard
I’m sorry to hear that. Having a good manager is so so so important honestly. One can be working on the hottest technology but if the manager is not good, the experience is ruined. Thanks for sharing your experience.
@@SundasKhalid Like you mentioned in the video, companies don’t understand what a data scientist is and because of that their expectations are x and my skills are y. It does hurt my performance review which I didn’t realize until after I started working as a data scientist. I’ve heard a lot of my friends who had bad performance reviews even though they had done a lot of work. But I just think companies have no idea what to do with a data scientist when they hire one.
@@jcdenton4281 exactly. My issue is that he asks me to do something, says no to my method, then does his way, it doesn’t work, and then we go back to using my way. Then blames me for taking too long to finish it. Every project without fail he’s done this.
@@boejiden7093 same all my thoughts are completely different from him. he likes to make simple model with complicated UI. who even cares about UI if model isn't doing good. and just two days ago they have to let me go because i am not fit for organization people like him are ruining data scientist posts and outcome of it.
After reading many of the comments, I've not heard anyone speak about the importance of working in a field that they love. When I grew up I first wanted to be an ornithologist, then a paleontologist, and then an astronomer. While I'm none of those now, except in an amateur role, I discovered that I really like the medical field. I did that by changing jobs frequently in my early years. The problems that I hear people complain about here are issues with corporate management. Many times there is nothing you can do about that. However, if you love what the company does in terms of a product or service, then that goes a very long way to mitigating the management issues.
Data Scientists sweat the details and this is a great asset for a product management. One casualty of being a data scientist is that you develop a expectation of higher proofs that many of your family and friends for everything. As a data scientist, I am fascinated not only by the data, but how the data was collected and obtained because understanding this often leads to more insights, but the ever present "well how do you know that is true?" question can often bump up against people who expect you to trust them at face value. I recently went back to university and the professor assigned books that had a lot of charts and tables and in typical form, I researched the data that was used and found that it wasn't up to par and while my professor agreed with me, I got the sense that she was perturbed that my discovery upset some of the foundational premises of her class. I ended up dropping the class because from that point, I checked everything and became obsessed with the data issues. If you are a person who likes being gullible -- not in a bad way, but a person who likes to believe in the good of others, think before becoming a data scientist.
Not all DS are like this though. In my team we have all sorts, with a few barely detail focused and more excited about the tech stack and automation. They do tend to come from software development.
thanks for sharing your experiences! The manager who spoke of wanted applied scientists (a category I wasn't aware of before this video) was odd to me. So he was saying that there wasn't enough data science related work for a full time hire (or else he wouldn't be wanting someone who could also serve as a software engineer). I can imagine companies having this need (smaller companies in particular where individuals have to wear multiple hats), but if I were a data scientist trying to get hired as a data scientist, I'd avoid this manager/company! I actually left a job in the 90s for this reason--I was being pulled into other jobs/tasks that weren't related to machine learning, so I left and joined a company I could do machine learning full time (I'm an old guy! I predated data science, so in that time, it was called "data mining" or "pattern recognition". and yes, I'm using machine learning and data science interchangeably even though I know they aren't fully interchangeable). The manager is key--even if you like the company, if you don't respect or are respected by your manager, it will be an awful job. Remember that when you are being interviewed, you are also interviewing them! :)
Thank you for sharing this. I thought I was one of the only ones struggling super hard with the ambiguities and absurdities of the data science job search.
The way you talk and laugh in between tells me you're free spirited. You got my kind of energy and i know you'd always be fun to be with. That aside, thanks for sharing this experiences as a lot are also trying to get a way around it.
You covered some really great points that I'm also experiencing in my current role as a "data scientist". Tbh I don't even know what I'm supposed to be doing and it makes me feel so lost especially since I'm still early in my career. I saw 2 great people leave the team to work as a data engineer and a software engineer. It's absolutely demotivating to see people leave. I'm also looking into transitioning into a more data engineer and/or product analytics role but the interview prep is super overwhelming/difficult and imposter syndrome really kicks in. 😢
I will try to help out here: 1. Descriptive analysis 2. Statistical inference 3. Machine learning 4. Causal inference These are the four categories that fall under DS role potential scope. Any business problem one could bring up should fit into one of these four categories
The key here too Sundas is remembering when it’s your turn to interview them. Recognizing what questions to ask, to see if THEY FIT YOU. I’ve been in many interviews where a recruiter was selling me into a role that I didn’t fit. A few, I ended, thanked the interviewer and told them “I’m not certain why we were paired up on this. I don’t have those skill sets, nor interest to work in that area”. That’s a REALLY HARD THING to do when your junior but, being aware and avoiding that pit fall is huge.
I am a data scientists & agree with everything what you have said. I found the video very insightful. I think the hard thing to sell is to convince the managers is that the success of any data science project depends on the 4 things. 1) data 2) computing 3) problem and 4) technical expertise or data science team. If any of the item is weak project will not work.
One of the biggest challenges I've had with being a data scientist is that DS/ML projects can be incredibly unpredictable. Managers always want detailed expectations and timelines, but when these managers have little to no DS background, these tend to become unrealistic. Me being a perpetual optimist gets me to buy into these expectations and timelines more than I should, which looks great to the managers at first, but unforeseen challenges with the data almost always hit me over the head eventually. As I build more experience I'll get better at managing this, but it's a common challenge right now.
I am learning Data Science from last 1 year, I had taken an online course and gave it full time means not done any job in that meanwhile. Now no one wants to hire a DS fresher with a career gap. All the internship and fresher job descriptions are scary to even look at. Companies are using buzz words like Data Analytics for Data entry jobs. Each and every job description is bombarded with various technologies which have career in themselves. Felling exhausted and frustrated from this whole Data Science thing. I know Data Science itself is an elegant career but due to the industry environment, I am learning Backend development now, Let's see what will come out of it.
During my last 3 years working in data science, I’ve always felt isolated. I haven’t had the opportunity to work with data teams, so it’s always been me as the only expert..
The reason why company’s don’t understand what a Data scientist is because it was a up job title by Facebooks lead software manger back in 2004. They wanted to hire a really smart mathematician to understand there data but he wouldn’t take the job because he didn’t like his Job title. So they told him he could pick his own job title.
I was placed as a tester and was asked to write test cases for a data transformation project. I worked hard, researched a lot, understood business requirements thoroughly and wrote over 100 test cases. Manager still gave negative feedback to my company and I’m so furious. I’m doing masters in data science and now working as a consultant DS but they just don’t understand what DS entails and keep referring me for any role that they get asked to fill. Next time I’ll make sure I’ll pick and choose my project as a consultant DS. If it’s not related to my skills I’m not taking the project. Even though I believe in agility, adapting to changing requirements and learning but some managers are not acceptable of growth mindset and have extremely high expectations !
" If it’s not related to my skills I’m not taking the project." Exactly. While it can be cool to take on a project that stretches your abilities, if you know what you are good at and what sets you apart, you should find those roles where that skill is appreciated. But all of us with that attitude had to learn it the hard way like you did -- so you're not alone.
You were definitely not wrong to be offended by your manager's comparison. I had a similar experience where my manager decided that only ppl who can make good data scientists are ppl with PhDs in physics, and he decided to only interview them for the role from that point on. I only have a MS, and it's in math, so whenever he brought it up how we need to make sure we get in "quality" people by having this filter, i got pretty upset
if i had a PHD in phyics... i wouldn't be working at a company like that lol id rather be a prof or working at a company that directly uses physics for the job (like nasa/etc) but never for data science
@@asdfafafdasfasdfs One of them had a PhD in Physics and the other MS in business. So i think the physics one decided he's the only qualified one, and the other guy was just impressed by the degree and went along with it
Thanks for sharing your experience, I'm embarking on the data science journey right now and it's good to hear some of the challenges so I can keep them in mind.
Preparing for interviews can be arduous because like you pointed out that different organisations have different definitions of a Data Scientist role and Data Science is such an extremely broad term that is often times disputed amongst people. You only find out when you sit for the interviews that the organisations sometimes are actually looking for a Data Analyst or ML Engineer or Data Engineer when the interview doesn't circle around what you'd imagine it to be. But that's the reality and thanks for pointing it out. And most of the time I am just scraping and cleaning the data which definitely gives me mixed feelings about my job role. I could really relate to your experiences and it definitely resonated with me. Such an insightful video for the ones who want to pursue their career in Data Science because often times it can be misinterpreted.
I have also worked in Data Science for 9 years now. I worked as a data scientist for my whole career until now. I was hired in direct as a Senior Software Engineer in Cyber Security. It's actually a really smooth transition.
@@Ricocase a lot more than what some might think. Think of this, code is data. Cyber security professionals want to eliminate exploitable code. Therefore, AI should help us find the vulnerabilities.
I am a recent graduate in Computer Science and want to instead go into Data Science industry instead of the traditional Software Developer/ Software Engineer career. So as someone like me who eventually one day wants to work in data science industry, is there an entry level job role that one can apply as a starting point and just get promoted/ level up to a data science in time rather than just trying to apply for a data science position from the get go? I dont mind starting at the bottom and working my way up the ladder. Anyone have any suggestions?
A long time ago, I experienced similar situations in other areas of data scientists and their relationships with managers, which required multidisciplinary work. I perceive from the comments that there is a communication gap between the data scientist and the manager. Once the problem is identified, sometimes it is time to make the route and perhaps make some diagrams to educate the manager on the most convenient alternatives to solve the problem.
I feel like tech jobs always morph: we have hiring expansions and more specialities, then we have layoffs and more generalizations. It does seem like the demand for data scientists dropped off a cliff, but I’ve been studying it a bit as a way to complement other skills. Thanks for the video ☺️
In Addition data scientists have to work pretty long hours, manage 3-4 projects and also have a poor work life health balance as compared to software engineers. Software engineers have more structured and stable lives in general. So the salary per hour is much higher for a software engineer.
A problem may be that a lot of companies are not far enough along on the data maturity model to actually use a Data Scientist effectively - so then they are surprised when they aren’t getting the results they wanted.
Totally agreed. Except for Big Tech, they don’t have the kind of user data or logging data that is ready made for current DS/ML tools. It takes more thinking to figure out a good way to use the data they have or figure out how to start collecting data that would be very useful. Maybe the Data Science role will also evolve. Most regular businesses will use the AutoML via chatGPT, and only those businesses that require maximum data knowledge will have Data Analyst/Scientist/Engineer/Strategy Consultant/Business Forecaster/Decision Maker etc. And it will be one giant field with many levels and sub fields.
I’ve been in the (fill in the blank) industry for ten years and never felt comfortable with the name “data scientist”. I never knew what to make of it. It just SOUNDS ambiguous lol To those on the struggle bus, I get it. My recommendation: I’ve always marketed myself as an analyst, highlighted the systems and languages I’m proficient in, and provided demo’s of past projects. Then, ask questions of your employer. Interviews are as much for them as they are for you.
My husband was referred to a data science role by a colleague and he was unsure about the role. It was hard for him to prepare because it seemed that people and organizations would use data analyst, data science, data engineer and applied scientist interchangeably. I am learning about it now since I have been wanting to get out of cyber security. Hope for the best.
I am not a data scientist but have worked with some. Thank you for sharing the harsh but truthful reality of Data Science jobs. Another harsh reality is ego-boosted people Data Scientists have to work with. When a PI gets funding big enough that they can hire a Data Scientist to look at their data or design their experiment, they are on a high horse and have already made up their mind about the results they must get. For example, a PI once was so sure that he could get Nature-level paper out of the results he had that a Data Scientist had to show him that randomly generated results with some tweaks had the same properties as those he was seeing in his result to burst his bubble. Obviously they never worked again together.
Hello! Love your channel. I'm a Sr. Data Analyst. You are spot on! when it comes to the interview process. I interviewed for a Sr. Data analyst role. I didnt apply at that point for Data Scientist bc I felt I was not qualified for it yet. One of my interviews for Sr. Data analyst was super technical and the interview was about ML, algorithms and AI. I had maybe 2 projects under my belt on that, so I didnt know all the answers and my take home assignment was ok, but I failed presentation questions. I felt discouraged. Then I had interviews at other companies for the same role and the questions were more about excel, sql, data bases, and some python. So yeah every company has different views on the roles. So the interviews can be too difficult for the "title" or too easy. You never know!
The 'harsh' reality of being a data analyst is that its really just 1 skill in a toolbox a good programmer has. That along with ML, microservices, event based programming.
I'm curious to know actually what a data scientist is. I'm a data analyst and my manager wants to hire a data scientist to automatize some of our data entry. It's a program that he doesn't understand so he wants to hire someone to run it. The weird thing about it is, we're a small dept within the city and only really work with data that pertains to us so I really don't see the point. I feel like half of our workload would be gone basically and everyone will be staring at walls.
When you get to a Senior level, things will become more clear. The low level work gets assigned to level 1,2,3. So when you really add value to the business and can impressively demonstrate your abilities regularly that's all that matters. These complaints about the DS role are all low level issues IMO.
I have been a junior data science for a week now, this week was filled with a lot of learning and frustrations. My first assignment was to analyze and evaluate the results of a computer vision model and its classification model. PS: This is my first time manipulating images.
This is a very interesting perspective. Not all companies treat data scientist alike. The role of a data scientist is an amalgam of skills from various backgrounds. Unless you start in a brand new team with the same background , the feelings of incompetency and being lost when you work for a traditional company are unavoidable. The way to deal with uncertainties and fear, is to continuously learn and improve yourself in the areas you feel less confidence. Thanks for putting these thoughts out!
I have been working in this area for over 25 years (way back before the data science tech tree). Companies are still trying to grapple with this domain and I tend to stay away from companies that don't really know how to use data science. My recommendation would be to stick with companies where the core technology is based on data science/AI. E.g., speech recognition, however, this gets to be more applied science work and usually requires a bit of training in the specific domain.
I think this is the paradox: companies that uses and understand the topic appreciate the work, but yet have higher expectations on the offered skills. Because they have also an established team in place. In contrary, companies that are merely in the beginning utilizing it having a hard time to define what skills they need, how they trying to achieve it, and fuzzy about the work you must then do. Yes, avoid them if your already at senior-level.
This was an enlightening video with great insight and I really enjoyed listening to someone articulate the challenges of a data science job when an individual transitions to a "Non Data Science" manager.
This is how I feel, first you work 2/3 or more of your life, so I believe that you should be happy working at whatever job you want. I,m 75 and last year just started programming with Python, got a long way to go though, but Data Science does sound interesting. I pray that you will find the meaning of your work and not let someone decide it for you. Lots of luck.
Can you please make a video for those unsure between Software Engineering/development and Data Science who are just learning to code(transitioning from a non-techncial carereer)/What steps should be takend for each path to get to an entry level role and if Switching between the two paths is common/easy?
@@binyaminkhawaja1536 Let me help this way, both roles deals with code and data. Just that the software engineer role is code heavy while the data scientist role is data heavy so it depends on what you are passionate about or like more, I don't know if this helps.
I spent the last 2 years learning data science while applying everywhere to get my foot in the door what surprised me was just how many roles there are out there that require data science/analytics skills but aren't themselves data analyst roles, they could be logistics/managerial roles at a warehouse, I've seen loss prevention roles need data skills, sometimes they turn out to be nothing more than data entry roles, quality control roles, etc etc and then of course the actual data roles themselves swing considerably, sometimes it's a business/financial analyst role and you're just pulling down charts and graphs to put on reports for the big wigs, and sometimes it's actually a full stack dev role that happens to include SQL and R in its language requirements, oh yeah and DB administrator roles which turn out to be IT helpdesk roles lol
I totally agree with you. The hardest part for me is to deal with those "data scientists" who barely scratch the surface of this world. I think that this stems from the problem that data science actually begins where one usefully combines the domain expertise, analytical data understanding, mathematical approaches, algorithms and all those software tools. There is soo much to learn, it will take you about 5 years fulltime on top of a basic MINT subject until one really knows what we are doing. Many beginners don't see this and can't understand it how hard this actually is. And the payment isn't worth it in most cases, because your managers won't understand this either.
I left the whole career line before I even really began! I studied Business Analytics for my masters with data science intentions in mind. After over five hundred application rejections (yes a real number - I'm an international graduate which makes competition way more intense) for jobs that are actually (looking at the requirements) obviously intermediate but for some reason termed as entry level, I decided to forget about this whole line altogether and focus on actual software development. Now working as a Unity game developer, and with my strong frontend development skills planning on taking on a frontend dev job next. At this stage I'm not interested in the least in anything that has to do the data science and it will probably stay this way.
@@milanacharya4865 i learned data science but didn't get a job because just after i finished my course the lockdown happened so after the lockdown i joined as a javascript devloper and learned nodejs amd reacts ...i am again thinking to switch in data science....so should i stay in MERN stack or should i move to data science...please reply as soon as you see it
I'm in my 50s and the only thing I can add is after 20 or 30 years the job becomes totally boring but you are stuck in the job because of the pension. I would advise younger data professionals to increase your skill set and transition out of the field because after several decades, its a completely boring job yet you still have to keep on top of latest developments, read papers, attend conferences etc when in truth you are really not interested in the job and you just want to go home, tend to your garden or finish that hobby project etc.. Roll on retirement.
Lots of posers out that who don't realize how hard it really is and also lots of people making it seem like its the easiest job in the world to do... "Oh, you just need to learn some basic programming, math, and stats," I hear this all the time. Definitely the ambiguity of the role also. Many companies don't even know what a Data Scientist is, don't have a hiring strategy, and/or don't know what they will do with this person once they are hired.
I left my data science career path and embarked on my actuarial path. I think the future of data science will be a skill rather than a career. Although data science is a great career, for me I realized it's the domain knowledge that it currently lacks that made me get a job as an actuarial analyst rather than as an actuarial data scientist. I might be wrong. My advice would be, in doing what you love, combine data science skills which will make you more robust in your work deliverance.
You are 100% correct. Data science is more of a skill than a career. I am a fullstack software engineer but I am currently running two Master's degree program in Data science just to get the knowledge and apply is to my Software development career.
I honestly think "data scientist" is a job title that is too young yet, HR people usually will take a couple of years (or decades) to wrap their heads around what the job entails and what kind of people they're dealing with. I mean it's been 40 years and they are only now beginning to grasp the role of software developer ^^
People are moving into project management because it is more concrete than data analysis. Another aspect is that people enjoy working within a domain. Focusing on a product domain lets someone stay in a workflow that gives them interesting domain knowledge.
Agreed that managers wont appreciate your usability. I urge you to guide us how can a data scientist work as a freelancer or a individual entrepreneur. In essence, elevate him/her from an employee / subordinate position.
My impressions of the industry is that there is so much skills/experience overlap , and companies look for multi-faceted workers, that there is some confusion in the hiring and project phases. One confusing aspect of this industry (Data Science) is not everyone has to be educated in Computer Science ( with degrees). So that leaves a lot of open-ended questions sometimes I think. I do not have a computer science degree (but I do have BS and MS STEM degrees). So I was surprised when I found out I could make entry into this field ( I am highly analytical and love math). But like you said companies want "fungible" ie, multi-facted workers SOMETIMES, not always. I can understand that.
I was trained as a statistical analyst in what might be called classical regression, DOE, and inference. At work, I analyzed data and leaned as much as I could about what is now referred to a data engineering. But, it was extra. Now, a data scientist is more like computer science or data management than statistics. I think that executives drove the push from data visualization because there was always antagonism with control versus the need for specialists in an esoteric field. There is still a need for understanding of the classical methods, yet most companies outside of clinical trials have jumped on the AI/ML bandwagon. Add the increasing chasm between management and staff, the vagueness of the term data scientist, and this gartner stuff and it seems like instead of simplifying and harmonizing the field per Jeff Wu, it seems oppositely like total anarchy, fad chasing, and terminology babble to me.
hello mam, love from India I started my data science journey from last 3 months. I successfully completed python,pandas,numpy,and sql right now i am lerning statistics. i always come to your channel whenever i need some clarification on data science and i also followed you on linkedin . thank you so much for making such amazing videos.
My company hires a lot of data scientists, and then tries to turn them into data analysts. Why? because the hiring managers are data analysts. and are too lazy to learn even the basic fundamentals of data science and machine learning. Hence, why we have almost 100% turnover in data science. usually, every 6 months.
Thank you very much for sharing your experiences! I have a PhD in Biomedical Engineering & I can honestly say that I've experienced many of the same experiences with interviewing & jobs just not understanding of defining your roles well. I was considering transitioning to Data Science but maybe I will look more into data analytics instead. I'm burnt out at this stage & just don't have it in me to take more chances.
Hi Sundas ! Thanks for sharing the experience. I am a new Machine Learning Engineer in the industry. I think I can resonate a lot with what you said, but at the same time I believe that the industry is getting more mature in terms of the term "Data Scientist", a lot of companies are getting more process and experience driven than nomenclature driven. People spending a lot of time at a single organisation and not moving across or not communicating with others in the industry may lead to fixated mindsets and definitions about Data Scientist, Applied Scientist etc. where all these definitions can vary a lot in the industry. Also, I am fortunate enough to join the industry (especially my company) at the right time, where it starting maturing, started having understanding equal to (if not more than) engineering processes. Also, for the people who want to join the industry now can get better guidance, from more mature industry personnel and processes. (Although, I doubt if FAANG has still learned that ;))
More videos should mention that having a computer science degree and domain experience are big benefits, not essential, to working in this Data Science industry. So if you are weak in one or two areas of Data Science, then having a CS degree or certifications, and domain experience, are very helpful .
Sundas, you are right that we should judge if the role is good for us from the interview questions. We shouldn't forget that the people they hire before us went through the same questions. Do you want to work with a team of Data Scientist that only know what they are asking in the job interview?
Well there are times when some statistical studies are used in the early phases of Product Design to discover what consumers or customers mostly desire in the product concerning it's function. Then it goes to the Engineering Teams M&E to Design it in. I see quite a few of them now.
As an entry level role for those entering in the field. Which path of these 3 roles would you say is best to begin and progress through... Data Engineer/Data Scientist/ Data Analyst? I've studied each and doing projects but equally like each as they somewhat overlap. What is more realistic junior opportunity to get hired in first?
@@movievibes521 but i hear somewhere to do first: DE then DA and then DS. Because DE is capable of doing all things of DA and DS. But just he is very busy. So DA AND DS take care of the rest. Plz explain your point of view. Thanks!
I have a background in mechanical engineering and worked in an aerospace manufacturing firm and it wasn't nearly as exciting or sexy as they made it sound. I think this is a problem with a lot of technical positions. Most of the work is quite repetitive and boring. I currently work as an industrial / Continuous improvement engineer, but have absorbed data science responsibilities. I would agree that most of the work is typically just cleaning data. Mine might be a little more interesting with a continous improvement background because I get to lead project initiatives and lead problem solving events across the company in addition to data science work. But overall I think its just companies trying to market themsevles to hire employees when most of the work in quite mundane.
This is soooo spot on and this made me feel so much better. I went on so many long interview process and then they would just ask me “SQL” questions..I was baffled..
I've given up trying to get a data science job. My goal is to become a data analyst/reporting analyst (Tableau/Power BI). Only pays about 20% less but less than 50% of the stress in my opinion
I guess I'm an 'old head' as I got into the field back in 2005 but the worst part is the fact they interview for deep learning and leet code and other things the hiring team doesn't really understand and then you get the job and you're just slinging SQL because there's no infrastructure, the data is a hot mess, and the culture is not at all ready to be data-driven. There's a reason 80 - 90% of data science/machine learning/AI initiatives fail (per Gartner and the MIT Sloan Management Review) and that's because the initiatives are set up by people still on the hype curve then sent out into the world with no actual plan. So data scientists are set up to fail and sure enough, they get the heck out of the industry, often within two to three years.
In general in IT if you have good communication skill than everything become secondary, you can take leadership track and manage bunch of people who are great coders but cant talk well and prefer to deliver quietly. We call them coding monkeys 😬, just tell them what to do and they will code happily and come back for more when done. In long run, its the leadership that brings money, fame and success and not the stand alone work.
And we call managers who know nothing about the product they're managing, trash managers because they will never climb the ladder using jargon and empty platitudes
I did not understand a word of this. How can a manager misunderstand what your job is. You either have the skills and qualifications to do it or not. If you have the skills to do it as a Data Scientist, how and why is it not part of your job? In my head if you need to work with data, it is data science. Are you expected to work on the front-end, back-end or what are we actually talking about here? Specifics please.
Whether you are a data scientist, mech engineer, etc., you should want to do that role because you love the subject. You don't study those fields because you want to get a good job. You don't study those fields because you want to move up the ladder as you said. If you want to move up the corporate ladder, it would be wise to learn necessary business courses like marketing, finance, leadership skills, et al. Learning data science to move up the ladder is a poor career strategy. The purpose for working at a company immediately after school in any of these specialized is to pay off your student loans and put money in the bank. While you are doing that you should be learning how to develop your own business on the side in the field of study you love. That way you can just leave your corporate job and don't have to worry. There are a few red flags in your video about yourself. The fact that you are referring to your coworkers as family doesn't bode well. They are not your family, they are your coworkers. They have their own aspirations, which more than likely differ from your own. Your family are the kids you make and the man you marry. And considering you're 31 years old, the fact that you are calling your job 'family' tells me you have some misplaced maternal love.
With all due respect, I think you misunderstood. She's not saying that her co-workers are her family, she is saying "job family" as in "group of jobs". On another note, your family is not necessarily "the man you marry and the kids you have", that is for each person to decide for themselves ;)
What we refuse to accept is the reality, that Data science, analysis and engineering were not meant to be standalone roles in the organisation. You are supposed to use Data related jobs as complementary skillsets like Excel, Word, PowerPoint and other programs to your other main job description. You can be a medical professional and data analyst with medical data. As an Educator, data science, analysis etc complement my student data processing and visualization. HR personnel needs the Data science/analysis skillset as a complement not stand alone. Else you will find yourself in and out of your primary job functions.
I'm 51yrs old and have been working in the operating room for 26yrs. Name a surgery and I've done it. Not a surgeon, but I work as their surgical assistant. Their right hand man, so to speak. I want to change careers and get out of the operating room. Can you explain how or where someone my age can begin a new path to being part of a data analyst with medical data like you mentioned?
@@mikel6787 There are lots of medical data that can be analysed in the field of medical surgery. You'll have to figure out how to gather relevant data in surgery that can be analyzed to improve surgical practices, save patients and make you the goto analyst consultant for a fee.
Heya! I dream to be a data scientist but i'm bit confused what degree program I should take.I mean is it better to go for 4 years software engineer degree then do masters in data science or do 4 year degree in data science and apply for data science job?Please help
I believe there is a massive discrepancy in "the perception of a true Data Scientist role" between the industry and job candidates. When you use data to solve a business problem, it is data science. On a high level, DS can be categorized as analytics, ML + inference focus, and data engineering (this is What LinkedIn and Airbnb categorized). When you say "A true Data Scientist" in your above video, you are referencing "ML or casual inference focused" DS. However, the DS field has become more specific compared to maybe 10 years ago when only the ML-focused roles were called DS. When you say you got asked only SQL (and I assume product case questions, like a consulting interview), you refer to a "Data analyst" role. However, these are also called Data scientists with an analytics focus. The boundary between data analyst and analytics DS becomes blurred, and I will use seniority and scope of work to differentiate them. If you are solely searching for ML focused DS, you can apply for "Machine Learning Engineer" or DS with an emphasis on modeling in their job description. I do have to mention that the (Product) Analytics-focused DS has a very large demand. Meta, LinkedIn, Airbnb, DoorDash, Uber, Spotify.. all these big names are hiring for these positions.
hello. I read your comments. what would you suggest for someone 29 years old lawyer (masters) who just wants to get into the field? would u suggest getting into data science and working for law firms? analyzing data to help the firm grow?
hello. I read your comments. what would you suggest for someone 29 years old lawyer (masters) who just wants to get into the field? would u suggest getting into data science and working for law firms? analyzing data to help the firm grow?
One more serious reason not to join DS (and to become Salesforce dev) is that managers enforce to find correlations in datasets where there is no correlation. In such a cases data scientists find out that "NO is also the answer" only in science, not in business. I saw one data scientists completely in depression because his manager insisted that after the advertising campaign there MUST BE a surge in profits, but there was no surge, and the manager insisted that this data scientist is just a bad professional. I also attended a lecture by the chief data scientist of Atlassian, where he told how for years he was bringing statistics to his managers that they did not like. In other words, data scientists often have to bring BAD NEWS to their superiors, and they REALLY don't like!
To be fair you said it yourself already that data science actually covers a family of domains or jobs. But I think indeed DS job is still too new and really rides on a lot of hypes and so a lot of the companies just didn't want to be left behind yet at the same time didn't really know what they're doing themselves. Also in reality there is yet to see that many companies (or even industries) truly benefit from having a DS or even just ML team from the business standpoint or profitability point.
All data scientists with RUclips channels explaining all carrer's pros and cons are doing god's work for us that are trying to break into the field. Thank you for your time and shared experiences.
I just left a DS role and took another job as a data engineer. Lots of companies think they need DS, but aren't quite sure what to do with the DS they have on staff. Personally, I ended up doing a lot of ad hoc, proof-of-concept projects. But most of my workday involved assembling data in SQL for domain experts to review. When you're the only guy on the team who knows SQL, SQL is all you'll do :)
@@HarshvardhanKanthode Domain expert != coding and data skills. Most of the people I was working with were engineers with deep product knowledge, but without coding and data skills.
As an outsider to the software/tech world, I am envious that discussing mental health, toxic work environments, mental happiness is even a thing. In every other industry (except TV news and mainstream news media) if you brought up those terms people would first look at you like you were an alien from Mars. Then everyone would start laughing at you for being such a soft woosie.
i think the data scientist -> product manager thing is a natural progression for people with quantitative skills who are also interested in entrepreneurship. they start out by wanting to apply math to business and then realize how much more interesting the business side of things really is.
I completely agree to what you say , the issue is the existing product manager are unable to think datascience solutions for the product. I have seen a lot of products where they should consider AI features primary they end up treating those as secondary. This makes their product outdated. I understand what you said about applied scientist, but its a hard fact.
I hear ya. I saw a lot of job posting recently with title Software Engineer but job description is more on the data analyst side. I recently applied and interviewed for a Software Engineer position and turned out during the interview, the hiring manager told me he wasn't sure why programming languages are in the description, but his team only deal with SQL and data like 90%.
The whole applied science role seems pretty typical for tech, sounds like the DS version of full-stack. Get more done with less, treat people like resources...not people etc. Everything you said seems accurate and everyone has at least one bad experience in tech. I think everyone in tech should have a back up plan/exit plan when they are entering in to it. Either moving up to management or something different.
It's because managers continuously work with data. Data is a measurement of how a product or service performs so keeping an eye on data is very important.
Nice! I see you use the Momentum extension in chrome. I've used it for years. Such a good extension. Soo thank you for this video. You gave some good potential warnings. I am completing the Google Data Analytics cert and think it might help me in my current job. I planned on taking a Data Science cert after but I'm not sure. It depends on how things go but perhaps I still will. I'm confident I'm capable, even after listening to your points. But I think I have a better expectation now. Thank you :)
I preioct that applied scientists are just going to become the new data scientists. Most companies that aren't FAANG level expect you to have a somewhat broad field of expertise. If your team isn't huge, you're expected to take care of many different kinds of projects that come up.
It was like I was listening to my own story. My dumb manager with a software engineering background never understood data and data science. And then he totally pivoted the team's purpose where I was left underutilized. However, I got a good role after all this happened, but the transition time was unimaginably killing and stressful.
The role of data scientists in my company is quite ambiguous. My job as a data scientist is more like a tester. Some of my colleagues whose job title is also data scientist are more like developers. I spent most of my time studying business logic, looking for something inappropriate in the current system, and trying to find a better way to build analytic models. Some of my colleagues spend most of their time realizing the model, tuning parameters, and refining models. If we are going to turn the analysis into a product then I do most of the design as well as quality control work like a product manager and my colleagues do most of the implementation work. I guess it's more appropriate to give us different titles, like data analyst? scientist? engineer? I need to turn to a dictionary to tell the difference lol.
I love being a Data Scientist and have never had a better job. To me the job of a Data Scientist is to translate a business problem into a math problem which can be solved with the available data. Solving it is the easy part. To do this, we are part business analyst, part financial analyst, part data engineer, part software developer, part machine learning engineer, part statistician, part visualization engineer and part product manager. The ambiguity is off the chart and the need to learn is constant. When I hire, I look for determination, intellectual curiosity, desire to constantly learn as well as the ability to deal with ambiguity and failure. The hard skills I can teach.
Hello Rick, it's an interesting job based on your explanation. I work as a logistic business analyst and I've been trying to switch to data science and really need a good mentor to set my mindset in place. Could we connect? I'm really looking forward to hear your feedback 🙏🏻
To me y'all are half statistician and half analyst. From the outside looking in as a data analyst.
You're no data engineer.
I love this comment. As a data scientist, curiosity, enthusiasm is the key. For me 2 years now freelancing in the field and I know how it feels. Another important part, you must be a team player and in most cases have a good managerial skill in order to advance in the career. I wish I get the opportunity to work with you someday.
Hey is masters important for this job role if you wanna get into big companies and mncs? Or should I just complete my undergrad?
I'm learning Data Science and whenever I read job descriptions for entry levels or even for Interns they want you to be jack of all trades and on top of that require you to have 2,3 years of experience. Plus they include big data streamlines and ask you to be an expert on it too along with all sorts of cloud services. I mean it feels like they're demanding an Intermediate Data Scientist and terming it as "Data Analysis Intern" or "Junior Data Analyst" role. This frustrates me a lot.
Me too. I feel you 😢
@@weinjien5436 do you think that would be enough? I need to make one of those courses.
@Raxosc 14 online course will be very helpful to prove ur skill sets (given that you dont have any experience)
@@weinjien5436 For a minute there I thought you said "14" online courses. As for me I know GCP basics.
@@123arskas 🤣🤣🤣🤣
In my personal experience the harsh reality also included:
1. Studying all the time, to the point that you personal life gets affected. 8 hours of work and 4 hours a day of just keep up with the new lake, cloud , software, technique, library , bi tool etc.
2. The feeling of "You don't know enough", "You are not enough". You briefly mention this and I am glad that I am not the only feeling the same way. I know python enough to do my job, but I am not a python expert who can write functions from scratch, or change between languages say from python to java or R. I use the existing libraries and move on to the next project.
My question is, would you do this kinda thing if you weren't getting paid?
You don’t need to keep up with the new stuff. It’s not like companies have unlimited money to switch cloud when a product releases. Or change a whole working thing because this bi tool is cool. Tf you work
This is exactly why I love being a Data Science... I'm always learning and it never gets boring.
same i feel as full stack developer
Build foundation. If you understand the basic concepts of object oriented programming, basic computation, os (unix based kernels), then keeping up with trends isn't all that difficult. Technologies, libraries, frameworks, etc. are an abstraction of the fundamentals.
You'll never "know" enough. It's not about knowing. It's about collaboration, researching the issue, and gathering the best tools to solve the problem.
Its really difficult to work as a Data scientist because every company have different expections from Data Scientist. And now a days, a data scientist should know all the skills like computer vision, deep learning, Operational research, Machine learning, SQL and many more. There must a discrete line between data scientist work roles. And also as you have mentioned interview prep, it is really difficult because we need to prep from scratch. - Struggle of a DataScientist
But many online institutes provide training saying anyone can be data scientist and get placement too.. with good CTC of minimum 7 L is that true,
Do any one transist this career in Data scientist ? Is good option ?
I agree with the part about the necessity of being versatile. Of course, looking for a job, one can just ignore the job offers containing techs one does not know. But in my view, there are so many of them that after that 'offer elimination' process, one's final list could turn out to be pretty short. Right now, I work as a Mid Data Scientist, and although I feel confident in what I am doing on a daily basis, I have (and constantly update) my 'to learn' list.
I don't know whether to laugh or cry when you say you are expected to know Operational research as well as the other stuff. Operational Research is huge (not just optimisation). It certainly requires data analysis and statistics skills - and SQL and communications and modelling and data cleaning..... I found that people are in awe of the trivial stuff that they can almost understand (e.g. spreadsheets with actual formulae!) and the challenging stuff is so far over their heads they don't even realise it is needed or even that it is there. I can relate to Sundas Khalid when I think of managers I had who really did not appreciate what the discipline offers. I had to do the stuff they asked for while trying to slip in the stuff that I knew would be useful. Now I can see that the same applies to related topics such as Data Science.
Ok I thought I was crazy feeling like I had to learn financial risk modeling from scratch before my interview and do a little quick project to show my ability lol This being for a Quant risk analyst role at a bank
@@OMPRAKASH-tl6yc hi
People usually leave managers, not jobs. Your experience there is not unusual. I also work in IT, and have resisted moving up to management. I'd rather be involved with the work, rather than stepping back to manage the workers. However, this has hurt my career in some work environments (Notably business, less research) where not wanting to move up into management is seen as a lack of motivation. So, for me, I've taken hits in compensation in order to support my own job satisfaction. I'm now old enough that retirement is on the horizon. I look at the size of my nest egg and contrast it with the ulcers that would have been required to make it larger, and I don't have many regrets.
So few people get this. It's not just about getting ulcers, but avoiding getting promoted to a level of incompetence. Not to offend anyone, but to be a manager is a completely different skill set that even many of those pursuing it don't fully grasp. Not to mention how even great managers aren't able to act beyond what they're allowed to do by, let's say, poor company policy.
This is why there are so many inadequate managers present in so many workplaces across every possible industry imaginable.
A good, great or awesome worker doesn't necessarily make a good manager. Also, being a manager isn't always more valuable than a skillful worker. (In any field)
Thus, it's a shame there are examples such as yours where a person takes a hit in compensation or any other benefit for what is a silly reason.
Have you ever tried to be a manager?
Yes there’s ambiguity in role definitions. In my previous company, I worked as Data scientist. However I hardly did any data science work. It was more like a developer work where I was assigned to test and deploy the ML models in production. I could learn a lot about aws cloud architecture and services. Trouble shoot real time issues. But did little of ML. Then I changed my job and joined as Data analyst. Here I do little of analysis, SQL etc and more of ML like developing models and fine tuning them. Cleaning the data, feature engineering and feature selection.
I would caution the people aspiring to enter the field because many a times companies themselves are not sure where and how they are going to use data science, what are the prerequisites for a successful data science project. They lack quality data, if they do have data- accessing and compiling the data is another issue. After aligning with multiple departments as single man army, you are expected to give some magical results. When results are satisfactory, you will be asked to deploy them. Here again the streamlined organisational requirements are missing. Somehow you drag the project. Later you are expected to monitor and troubleshoot issues in production. And then after all of this single man army efforts, management doesn’t see any value in it. Just to keep up with trend companies end up hiring a data scientist and expect that person to assume multiple roles and give them an end to end solution. They are just giving analytics a try without investing much in it.
Thank you for you honest assessment!
exactly facing the same here. How was the transition from developer work to actual ML like?
Thanks for sharing!
Thank you for sharing! I have a really similar story. Worked as a CV/ML for a 3 years and constantly tried to be up to date with last tech stack and ml frameworks but only few time had real chance to use them. Now I'm in process to switching to frontend.
@@dariashtanakova1200 front end developer will be completely different field right ... You will need more coding html, java etc ... What are other options you can opt for?
I'm also a data analyst, after completing college degree applyed several job, after got rejected from more then 50 companies. Got i job in one company with heavy workload plus nightshift too. Then i decided to leave it.
Then decided to change field.
And now i started e-commerce online business and put my all computer skills in business. Most of my work done by AI.
And i really happy with my decision
👍
Great insight! I got into data science right after college and really struggled to do so. The fact that the data science job role is defined so differently at different companies made interviews really hard to prep for. It felt like you need to be a jack of all trades and be well versed in data science, data engineering and data analyst principles and it was just a lot to prep for. Plus a lot of roles require graduate degrees. After I got a role, I left after a year and transitioned into SWE as like you said the expectations didn’t match the reality plus I was also at a toxic workplace. I realized that I want to be in a role that is more well-defined and somewhat uniform across companies so if i need to look for a job later, I’m not being restricted to only a subset of opportunities bc of the different definitions and requirements.
Hey! I am actually trying really hard to get into data science but there are a few complications that i am facing and I don’t have a degree in this field, i spoke with a few counsellors as well but they couldn’t provide me with the answers that i was looking for, in the end I believe someone who is already in this field can help me out with my queries, so i was hoping if you could like help me a little on a few concepts?
Hi thanks for insights.. it would be helpful if you can share your linkedin id..
Yeah! Totally agree. As a bioinformatician, I also felt some things this video reported. I think the "data scientist" role is even newer and less well-defined. So, managers and colleagues don't know what it is, which leads to great mismatches in terms of expectations...
As far as industries evolve into more data-centric approaches, I guess people will be better valued...
@@Baazkingdom ur number?? I want to talk to u
So basically, in the industry, billions of dollars or more a year is being wasted on faulty recruitment methods that waste interviewer time.
The biggest problem I’ve had so far is working with my manager. We both have very different perspectives on various problems but he simply doesn’t even want to hear my perspective even when my perspective is usually right. We are both data scientists but he graduated as a data scientist back in 2003. So he’s an old head and I recently graduated. It’s very tough to get along and having your voice heard
I’m sorry to hear that. Having a good manager is so so so important honestly. One can be working on the hottest technology but if the manager is not good, the experience is ruined. Thanks for sharing your experience.
@@SundasKhalid Like you mentioned in the video, companies don’t understand what a data scientist is and because of that their expectations are x and my skills are y. It does hurt my performance review which I didn’t realize until after I started working as a data scientist. I’ve heard a lot of my friends who had bad performance reviews even though they had done a lot of work. But I just think companies have no idea what to do with a data scientist when they hire one.
And we have to shut up cause of inexperienced argument
@@jcdenton4281 exactly. My issue is that he asks me to do something, says no to my method, then does his way, it doesn’t work, and then we go back to using my way. Then blames me for taking too long to finish it. Every project without fail he’s done this.
@@boejiden7093 same all my thoughts are completely different from him. he likes to make simple model with complicated UI. who even cares about UI if model isn't doing good. and just two days ago they have to let me go because i am not fit for organization people like him are ruining data scientist posts and outcome of it.
After reading many of the comments, I've not heard anyone speak about the importance of working in a field that they love. When I grew up I first wanted to be an ornithologist, then a paleontologist, and then an astronomer. While I'm none of those now, except in an amateur role, I discovered that I really like the medical field. I did that by changing jobs frequently in my early years. The problems that I hear people complain about here are issues with corporate management. Many times there is nothing you can do about that. However, if you love what the company does in terms of a product or service, then that goes a very long way to mitigating the management issues.
Data Scientists sweat the details and this is a great asset for a product management. One casualty of being a data scientist is that you develop a expectation of higher proofs that many of your family and friends for everything. As a data scientist, I am fascinated not only by the data, but how the data was collected and obtained because understanding this often leads to more insights, but the ever present "well how do you know that is true?" question can often bump up against people who expect you to trust them at face value.
I recently went back to university and the professor assigned books that had a lot of charts and tables and in typical form, I researched the data that was used and found that it wasn't up to par and while my professor agreed with me, I got the sense that she was perturbed that my discovery upset some of the foundational premises of her class. I ended up dropping the class because from that point, I checked everything and became obsessed with the data issues.
If you are a person who likes being gullible -- not in a bad way, but a person who likes to believe in the good of others, think before becoming a data scientist.
Not all DS are like this though. In my team we have all sorts, with a few barely detail focused and more excited about the tech stack and automation. They do tend to come from software development.
thanks for sharing your experiences! The manager who spoke of wanted applied scientists (a category I wasn't aware of before this video) was odd to me. So he was saying that there wasn't enough data science related work for a full time hire (or else he wouldn't be wanting someone who could also serve as a software engineer). I can imagine companies having this need (smaller companies in particular where individuals have to wear multiple hats), but if I were a data scientist trying to get hired as a data scientist, I'd avoid this manager/company! I actually left a job in the 90s for this reason--I was being pulled into other jobs/tasks that weren't related to machine learning, so I left and joined a company I could do machine learning full time (I'm an old guy! I predated data science, so in that time, it was called "data mining" or "pattern recognition". and yes, I'm using machine learning and data science interchangeably even though I know they aren't fully interchangeable).
The manager is key--even if you like the company, if you don't respect or are respected by your manager, it will be an awful job. Remember that when you are being interviewed, you are also interviewing them! :)
last line is just gold
@@neel6978 Old people are gold mines, I truly can't fathom how knowledgeable are old timers !
Thank you for sharing this. I thought I was one of the only ones struggling super hard with the ambiguities and absurdities of the data science job search.
The way you talk and laugh in between tells me you're free spirited. You got my kind of energy and i know you'd always be fun to be with. That aside, thanks for sharing this experiences as a lot are also trying to get a way around it.
You covered some really great points that I'm also experiencing in my current role as a "data scientist". Tbh I don't even know what I'm supposed to be doing and it makes me feel so lost especially since I'm still early in my career. I saw 2 great people leave the team to work as a data engineer and a software engineer. It's absolutely demotivating to see people leave. I'm also looking into transitioning into a more data engineer and/or product analytics role but the interview prep is super overwhelming/difficult and imposter syndrome really kicks in. 😢
I will try to help out here:
1. Descriptive analysis
2. Statistical inference
3. Machine learning
4. Causal inference
These are the four categories that fall under DS role potential scope. Any business problem one could bring up should fit into one of these four categories
The key here too Sundas is remembering when it’s your turn to interview them. Recognizing what questions to ask, to see if THEY FIT YOU. I’ve been in many interviews where a recruiter was selling me into a role that I didn’t fit. A few, I ended, thanked the interviewer and told them “I’m not certain why we were paired up on this. I don’t have those skill sets, nor interest to work in that area”. That’s a REALLY HARD THING to do when your junior but, being aware and avoiding that pit fall is huge.
I am a data scientists & agree with everything what you have said. I found the video very insightful. I think the hard thing to sell is to convince the managers is that the success of any data science project depends on the 4 things. 1) data 2) computing 3) problem and 4) technical expertise or data science team. If any of the item is weak project will not work.
Interesting. Never heard of an applied scientist before. Great content. Thanks
One of the biggest challenges I've had with being a data scientist is that DS/ML projects can be incredibly unpredictable. Managers always want detailed expectations and timelines, but when these managers have little to no DS background, these tend to become unrealistic. Me being a perpetual optimist gets me to buy into these expectations and timelines more than I should, which looks great to the managers at first, but unforeseen challenges with the data almost always hit me over the head eventually. As I build more experience I'll get better at managing this, but it's a common challenge right now.
I am learning Data Science from last 1 year, I had taken an online course and gave it full time means not done any job in that meanwhile. Now no one wants to hire a DS fresher with a career gap. All the internship and fresher job descriptions are scary to even look at. Companies are using buzz words like Data Analytics for Data entry jobs. Each and every job description is bombarded with various technologies which have career in themselves. Felling exhausted and frustrated from this whole Data Science thing. I know Data Science itself is an elegant career but due to the industry environment, I am learning Backend development now, Let's see what will come out of it.
During my last 3 years working in data science, I’ve always felt isolated. I haven’t had the opportunity to work with data teams, so it’s always been me as the only expert..
In a way I don’t mind being isolated from people but at the same time I wouldn’t want all the work load put on me
@@Mysterious.phanto what would you choose instead the DS with the knowledge you have now?
@@neel6978 probably software engineering
The reason why company’s don’t understand what a Data scientist is because it was a up job title by Facebooks lead software manger back in 2004. They wanted to hire a really smart mathematician to understand there data but he wouldn’t take the job because he didn’t like his Job title. So they told him he could pick his own job title.
Me: data scientist who is experienced with ml and dl.
Boss: let's solve this super complex modelling problem without using ml or dl.
I'm in the selection process for a Data Analyst course from a company, so thanks for shedding a light about this. Fungible content.
I was placed as a tester and was asked to write test cases for a data transformation project. I worked hard, researched a lot, understood business requirements thoroughly and wrote over 100 test cases. Manager still gave negative feedback to my company and I’m so furious. I’m doing masters in data science and now working as a consultant DS but they just don’t understand what DS entails and keep referring me for any role that they get asked to fill. Next time I’ll make sure I’ll pick and choose my project as a consultant DS. If it’s not related to my skills I’m not taking the project. Even though I believe in agility, adapting to changing requirements and learning but some managers are not acceptable of growth mindset and have extremely high expectations !
Not some managers. Most of managers
" If it’s not related to my skills I’m not taking the project." Exactly. While it can be cool to take on a project that stretches your abilities, if you know what you are good at and what sets you apart, you should find those roles where that skill is appreciated. But all of us with that attitude had to learn it the hard way like you did -- so you're not alone.
"I'm just cleaning data" this feeling is so true as a student. But they pay you well just to clean data lol.
You were definitely not wrong to be offended by your manager's comparison. I had a similar experience where my manager decided that only ppl who can make good data scientists are ppl with PhDs in physics, and he decided to only interview them for the role from that point on. I only have a MS, and it's in math, so whenever he brought it up how we need to make sure we get in "quality" people by having this filter, i got pretty upset
if i had a PHD in phyics... i wouldn't be working at a company like that lol id rather be a prof or working at a company that directly uses physics for the job (like nasa/etc) but never for data science
Out of curiosity, what was the background / qualifications of your manager?
@@asdfafafdasfasdfs One of them had a PhD in Physics and the other MS in business. So i think the physics one decided he's the only qualified one, and the other guy was just impressed by the degree and went along with it
Thanks for sharing your experience, I'm embarking on the data science journey right now and it's good to hear some of the challenges so I can keep them in mind.
Preparing for interviews can be arduous because like you pointed out that different organisations have different definitions of a Data Scientist role and Data Science is such an extremely broad term that is often times disputed amongst people. You only find out when you sit for the interviews that the organisations sometimes are actually looking for a Data Analyst or ML Engineer or Data Engineer when the interview doesn't circle around what you'd imagine it to be. But that's the reality and thanks for pointing it out.
And most of the time I am just scraping and cleaning the data which definitely gives me mixed feelings about my job role. I could really relate to your experiences and it definitely resonated with me. Such an insightful video for the ones who want to pursue their career in Data Science because often times it can be misinterpreted.
Basically HR don’t understand the difference between curtain jobs title so they’ll just slap anything on it
What is your salary pls tell
I have also worked in Data Science for 9 years now. I worked as a data scientist for my whole career until now. I was hired in direct as a Senior Software Engineer in Cyber Security. It's actually a really smooth transition.
What does data science have to do with cybersecurity?
@@Ricocase a lot more than what some might think. Think of this, code is data. Cyber security professionals want to eliminate exploitable code. Therefore, AI should help us find the vulnerabilities.
Pls tell what software engineer do
I am a recent graduate in Computer Science and want to instead go into Data Science industry instead of the traditional Software Developer/ Software Engineer career. So as someone like me who eventually one day wants to work in data science industry, is there an entry level job role that one can apply as a starting point and just get promoted/ level up to a data science in time rather than just trying to apply for a data science position from the get go? I dont mind starting at the bottom and working my way up the ladder. Anyone have any suggestions?
Your videos are a HUGE inspiration!! just started out my own youtube (from my experience as a data analyst) All the best!
A long time ago, I experienced similar situations in other areas of data scientists and their relationships with managers, which required multidisciplinary work. I perceive from the comments that there is a communication gap between the data scientist and the manager. Once the problem is identified, sometimes it is time to make the route and perhaps make some diagrams to educate the manager on the most convenient alternatives to solve the problem.
I feel like tech jobs always morph: we have hiring expansions and more specialities, then we have layoffs and more generalizations. It does seem like the demand for data scientists dropped off a cliff, but I’ve been studying it a bit as a way to complement other skills. Thanks for the video ☺️
In Addition data scientists have to work pretty long hours, manage 3-4 projects and also have a poor work life health balance as compared to software engineers. Software engineers have more structured and stable lives in general. So the salary per hour is much higher for a software engineer.
A problem may be that a lot of companies are not far enough along on the data maturity model to actually use a Data Scientist effectively - so then they are surprised when they aren’t getting the results they wanted.
Totally agreed. Except for Big Tech, they don’t have the kind of user data or logging data that is ready made for current DS/ML tools. It takes more thinking to figure out a good way to use the data they have or figure out how to start collecting data that would be very useful. Maybe the Data Science role will also evolve. Most regular businesses will use the AutoML via chatGPT, and only those businesses that require maximum data knowledge will have Data Analyst/Scientist/Engineer/Strategy Consultant/Business Forecaster/Decision Maker etc. And it will be one giant field with many levels and sub fields.
I’ve been in the (fill in the blank) industry for ten years and never felt comfortable with the name “data scientist”. I never knew what to make of it. It just SOUNDS ambiguous lol To those on the struggle bus, I get it. My recommendation: I’ve always marketed myself as an analyst, highlighted the systems and languages I’m proficient in, and provided demo’s of past projects. Then, ask questions of your employer. Interviews are as much for them as they are for you.
A sizable part of product / project management is data analysis / analytics, so its not unsurprising to see the easy transition.
My husband was referred to a data science role by a colleague and he was unsure about the role. It was hard for him to prepare because it seemed that people and organizations would use data analyst, data science, data engineer and applied scientist interchangeably. I am learning about it now since I have been wanting to get out of cyber security. Hope for the best.
Is cyber security not worth it for long term? I was thinking of pursuing this field.
I am not a data scientist but have worked with some. Thank you for sharing the harsh but truthful reality of Data Science jobs. Another harsh reality is ego-boosted people Data Scientists have to work with. When a PI gets funding big enough that they can hire a Data Scientist to look at their data or design their experiment, they are on a high horse and have already made up their mind about the results they must get. For example, a PI once was so sure that he could get Nature-level paper out of the results he had that a Data Scientist had to show him that randomly generated results with some tweaks had the same properties as those he was seeing in his result to burst his bubble. Obviously they never worked again together.
Hello! Love your channel. I'm a Sr. Data Analyst. You are spot on! when it comes to the interview process.
I interviewed for a Sr. Data analyst role. I didnt apply at that point for Data Scientist bc I felt I was not qualified for it yet. One of my interviews for Sr. Data analyst was super technical and the interview was about ML, algorithms and AI. I had maybe 2 projects under my belt on that, so I didnt know all the answers and my take home assignment was ok, but I failed presentation questions. I felt discouraged. Then I had interviews at other companies for the same role and the questions were more about excel, sql, data bases, and some python. So yeah every company has different views on the roles. So the interviews can be too difficult for the "title" or too easy. You never know!
The 'harsh' reality of being a data analyst is that its really just 1 skill in a toolbox a good programmer has. That along with ML, microservices, event based programming.
I'm curious to know actually what a data scientist is. I'm a data analyst and my manager wants to hire a data scientist to automatize some of our data entry. It's a program that he doesn't understand so he wants to hire someone to run it. The weird thing about it is, we're a small dept within the city and only really work with data that pertains to us so I really don't see the point. I feel like half of our workload would be gone basically and everyone will be staring at walls.
When you get to a Senior level, things will become more clear. The low level work gets assigned to level 1,2,3. So when you really add value to the business and can impressively demonstrate your abilities regularly that's all that matters. These complaints about the DS role are all low level issues IMO.
I have been a junior data science for a week now, this week was filled with a lot of learning and frustrations. My first assignment was to analyze and evaluate the results of a computer vision model and its classification model. PS: This is my first time manipulating images.
This is a very interesting perspective. Not all companies treat data scientist alike. The role of a data scientist is an amalgam of skills from various backgrounds. Unless you start in a brand new team with the same background , the feelings of incompetency and being lost when you work for a traditional company are unavoidable. The way to deal with uncertainties and fear, is to continuously learn and improve yourself in the areas you feel less confidence. Thanks for putting these thoughts out!
I have been working in this area for over 25 years (way back before the data science tech tree). Companies are still trying to grapple with this domain and I tend to stay away from companies that don't really know how to use data science. My recommendation would be to stick with companies where the core technology is based on data science/AI. E.g., speech recognition, however, this gets to be more applied science work and usually requires a bit of training in the specific domain.
I think this is the paradox: companies that uses and understand the topic appreciate the work, but yet have higher expectations on the offered skills. Because they have also an established team in place.
In contrary, companies that are merely in the beginning utilizing it having a hard time to define what skills they need, how they trying to achieve it, and fuzzy about the work you must then do. Yes, avoid them if your already at senior-level.
This was an enlightening video with great insight and I really enjoyed listening to someone articulate the challenges of a data science job when an individual transitions to a "Non Data Science" manager.
Thanks for the video. I was actually waiting for some DS to expose a bit of the harsh part of the career.
This is how I feel, first you work 2/3 or more of your life, so I believe that you should be happy working at whatever job you want. I,m 75 and last year just started programming with Python, got a long way to go though, but Data Science does sound interesting. I pray that you will find the meaning of your work and not let someone decide it for you. Lots of luck.
75?😳😳..you need to rest...why bother yourself learning
good advice...
😂😂😅
@@preciouseze1289 stimulating the brain especially at old age helps prevent brain pathology such as dementia
Thanks for watching! Do any of the things mentioned in this video resonate with you? What else would you add to the list?
Hey I am going to do data science in bachlor in 2023 is that right or I should choose other
Can you please make a video for those unsure between Software Engineering/development and Data Science who are just learning to code(transitioning from a non-techncial carereer)/What steps should be takend for each path to get to an entry level role and if Switching between the two paths is common/easy?
@@binyaminkhawaja1536 Let me help this way, both roles deals with code and data. Just that the software engineer role is code heavy while the data scientist role is data heavy so it depends on what you are passionate about or like more, I don't know if this helps.
I spent the last 2 years learning data science while applying everywhere to get my foot in the door
what surprised me was just how many roles there are out there that require data science/analytics skills but aren't themselves data analyst roles, they could be logistics/managerial roles at a warehouse, I've seen loss prevention roles need data skills, sometimes they turn out to be nothing more than data entry roles, quality control roles, etc etc
and then of course the actual data roles themselves swing considerably, sometimes it's a business/financial analyst role and you're just pulling down charts and graphs to put on reports for the big wigs, and sometimes it's actually a full stack dev role that happens to include SQL and R in its language requirements, oh yeah and DB administrator roles which turn out to be IT helpdesk roles lol
I totally agree with you. The hardest part for me is to deal with those "data scientists" who barely scratch the surface of this world. I think that this stems from the problem that data science actually begins where one usefully combines the domain expertise, analytical data understanding, mathematical approaches, algorithms and all those software tools. There is soo much to learn, it will take you about 5 years fulltime on top of a basic MINT subject until one really knows what we are doing. Many beginners don't see this and can't understand it how hard this actually is. And the payment isn't worth it in most cases, because your managers won't understand this either.
I left the whole career line before I even really began! I studied Business Analytics for my masters with data science intentions in mind. After over five hundred application rejections (yes a real number - I'm an international graduate which makes competition way more intense) for jobs that are actually (looking at the requirements) obviously intermediate but for some reason termed as entry level, I decided to forget about this whole line altogether and focus on actual software development. Now working as a Unity game developer, and with my strong frontend development skills planning on taking on a frontend dev job next. At this stage I'm not interested in the least in anything that has to do the data science and it will probably stay this way.
Hi Haritha, I have am in a similar situation like you were. Do you mind communicating with me regarding your journey? I would truly appreciate it.
@@milanacharya4865 i learned data science but didn't get a job because just after i finished my course the lockdown happened so after the lockdown i joined as a javascript devloper and learned nodejs amd reacts ...i am again thinking to switch in data science....so should i stay in MERN stack or should i move to data science...please reply as soon as you see it
Thank you so much for this. I thoroughly enjoyed your honest perspective and learned a lot. Keep at it!
Glad you enjoyed it!
I'm in my 50s and the only thing I can add is after 20 or 30 years the job becomes totally boring but you are stuck in the job because of the pension. I would advise younger data professionals to increase your skill set and transition out of the field because after several decades, its a completely boring job yet you still have to keep on top of latest developments, read papers, attend conferences etc when in truth you are really not interested in the job and you just want to go home, tend to your garden or finish that hobby project etc.. Roll on retirement.
Which career would you suggest us to transition in?
You're lying. If so where to "transition out of the field"?
Lots of posers out that who don't realize how hard it really is and also lots of people making it seem like its the easiest job in the world to do... "Oh, you just need to learn some basic programming, math, and stats," I hear this all the time. Definitely the ambiguity of the role also. Many companies don't even know what a Data Scientist is, don't have a hiring strategy, and/or don't know what they will do with this person once they are hired.
This is so insightful!!! Thank you for always being so authentic and an absolute queen all around. AMAZING video as always.
Thank you, Linda ❤️
I left my data science career path and embarked on my actuarial path. I think the future of data science will be a skill rather than a career. Although data science is a great career, for me I realized it's the domain knowledge that it currently lacks that made me get a job as an actuarial analyst rather than as an actuarial data scientist.
I might be wrong. My advice would be, in doing what you love, combine data science skills which will make you more robust in your work deliverance.
You are 100% correct. Data science is more of a skill than a career. I am a fullstack software engineer but I am currently running two Master's degree program in Data science just to get the knowledge and apply is to my Software development career.
I honestly think "data scientist" is a job title that is too young yet, HR people usually will take a couple of years (or decades) to wrap their heads around what the job entails and what kind of people they're dealing with. I mean it's been 40 years and they are only now beginning to grasp the role of software developer ^^
People are moving into project management because it is more concrete than data analysis. Another aspect is that people enjoy working within a domain. Focusing on a product domain lets someone stay in a workflow that gives them interesting domain knowledge.
Agreed that managers wont appreciate your usability. I urge you to guide us how can a data scientist work as a freelancer or a individual entrepreneur. In essence, elevate him/her from an employee / subordinate position.
My impressions of the industry is that there is so much skills/experience overlap , and companies look for multi-faceted workers, that there is some confusion in the hiring and project phases. One confusing aspect of this industry (Data Science) is not everyone has to be educated in Computer Science ( with degrees). So that leaves a lot of open-ended questions sometimes I think. I do not have a computer science degree (but I do have BS and MS STEM degrees). So I was surprised when I found out I could make entry into this field ( I am highly analytical and love math). But like you said companies want "fungible" ie, multi-facted workers SOMETIMES, not always. I can understand that.
I was trained as a statistical analyst in what might be called classical regression, DOE, and inference.
At work, I analyzed data and leaned as much as I could about what is now referred to a data engineering.
But, it was extra.
Now, a data scientist is more like computer science or data management than statistics.
I think that executives drove the push from data visualization because there was always antagonism with control versus the need for specialists in an esoteric field.
There is still a need for understanding of the classical methods, yet most companies outside of clinical trials have jumped on the AI/ML bandwagon.
Add the increasing chasm between management and staff, the vagueness of the term data scientist, and this gartner stuff and it seems like instead of simplifying and harmonizing the field per Jeff Wu, it seems oppositely like total anarchy, fad chasing, and terminology babble to me.
hello mam, love from India I started my data science journey from last 3 months. I successfully completed python,pandas,numpy,and sql right now i am lerning statistics. i always come to your channel whenever i need some clarification on data science and i also followed you on linkedin . thank you so much for making such amazing videos.
Where r u from
@@abhishekjadhav4105 uttar pradesh
How i start my journey as a data scientist from another field? what's the basic knowledge i have to start with
@@mirthplay4660 RUclips, udemy can help you.
How r uh learning data science
My company hires a lot of data scientists, and then tries to turn them into data analysts. Why? because the hiring managers are data analysts. and are too lazy to learn even the basic fundamentals of data science and machine learning. Hence, why we have almost 100% turnover in data science. usually, every 6 months.
This is exactly it.
This is exactly it. "We don't really do that here" was the response I got at my company when I expressed interest
Thank you very much for sharing your experiences! I have a PhD in Biomedical Engineering & I can honestly say that I've experienced many of the same experiences with interviewing & jobs just not understanding of defining your roles well. I was considering transitioning to Data Science but maybe I will look more into data analytics instead. I'm burnt out at this stage & just don't have it in me to take more chances.
Hi Sundas ! Thanks for sharing the experience. I am a new Machine Learning Engineer in the industry. I think I can resonate a lot with what you said, but at the same time I believe that the industry is getting more mature in terms of the term "Data Scientist", a lot of companies are getting more process and experience driven than nomenclature driven. People spending a lot of time at a single organisation and not moving across or not communicating with others in the industry may lead to fixated mindsets and definitions about Data Scientist, Applied Scientist etc. where all these definitions can vary a lot in the industry. Also, I am fortunate enough to join the industry (especially my company) at the right time, where it starting maturing, started having understanding equal to (if not more than) engineering processes.
Also, for the people who want to join the industry now can get better guidance, from more mature industry personnel and processes. (Although, I doubt if FAANG has still learned that ;))
Shubhankar sir please share ur number
More videos should mention that having a computer science degree and domain experience are big benefits, not essential, to working in this Data Science industry. So if you are weak in one or two areas of Data Science, then having a CS degree or certifications, and domain experience, are very helpful .
Sundas, you are right that we should judge if the role is good for us from the interview questions. We shouldn't forget that the people they hire before us went through the same questions. Do you want to work with a team of Data Scientist that only know what they are asking in the job interview?
Well there are times when some statistical studies are used in the early phases of Product Design to discover what consumers or customers mostly desire in the product concerning it's function. Then it goes to the Engineering Teams M&E to Design it in. I see quite a few of them now.
As an entry level role for those entering in the field. Which path of these 3 roles would you say is best to begin and progress through... Data Engineer/Data Scientist/ Data Analyst? I've studied each and doing projects but equally like each as they somewhat overlap. What is more realistic junior opportunity to get hired in first?
my suggestion would be da-de-ds bcz it would be incremental learning and you will be have strong base as data scientist
@@movievibes521 but i hear somewhere to do first: DE then DA and then DS. Because DE is capable of doing all things of DA and DS. But just he is very busy. So DA AND DS take care of the rest. Plz explain your point of view. Thanks!
I have a background in mechanical engineering and worked in an aerospace manufacturing firm and it wasn't nearly as exciting or sexy as they made it sound. I think this is a problem with a lot of technical positions. Most of the work is quite repetitive and boring. I currently work as an industrial / Continuous improvement engineer, but have absorbed data science responsibilities. I would agree that most of the work is typically just cleaning data. Mine might be a little more interesting with a continous improvement background because I get to lead project initiatives and lead problem solving events across the company in addition to data science work. But overall I think its just companies trying to market themsevles to hire employees when most of the work in quite mundane.
This is soooo spot on and this made me feel so much better. I went on so many long interview process and then they would just ask me “SQL” questions..I was baffled..
I've given up trying to get a data science job. My goal is to become a data analyst/reporting analyst (Tableau/Power BI). Only pays about 20% less but less than 50% of the stress in my opinion
time to change my ambition
I guess I'm an 'old head' as I got into the field back in 2005 but the worst part is the fact they interview for deep learning and leet code and other things the hiring team doesn't really understand and then you get the job and you're just slinging SQL because there's no infrastructure, the data is a hot mess, and the culture is not at all ready to be data-driven. There's a reason 80 - 90% of data science/machine learning/AI initiatives fail (per Gartner and the MIT Sloan Management Review) and that's because the initiatives are set up by people still on the hype curve then sent out into the world with no actual plan. So data scientists are set up to fail and sure enough, they get the heck out of the industry, often within two to three years.
In general in IT if you have good communication skill than everything become secondary, you can take leadership track and manage bunch of people who are great coders but cant talk well and prefer to deliver quietly. We call them coding monkeys 😬, just tell them what to do and they will code happily and come back for more when done.
In long run, its the leadership that brings money, fame and success and not the stand alone work.
its people like them who dont have patience and analytical skill to do coding and only knows to talk. sadly they are paid more
And we call managers who know nothing about the product they're managing, trash managers because they will never climb the ladder using jargon and empty platitudes
I did not understand a word of this. How can a manager misunderstand what your job is. You either have the skills and qualifications to do it or not. If you have the skills to do it as a Data Scientist, how and why is it not part of your job? In my head if you need to work with data, it is data science. Are you expected to work on the front-end, back-end or what are we actually talking about here? Specifics please.
Whether you are a data scientist, mech engineer, etc., you should want to do that role because you love the subject. You don't study those fields because you want to get a good job. You don't study those fields because you want to move up the ladder as you said. If you want to move up the corporate ladder, it would be wise to learn necessary business courses like marketing, finance, leadership skills, et al. Learning data science to move up the ladder is a poor career strategy. The purpose for working at a company immediately after school in any of these specialized is to pay off your student loans and put money in the bank. While you are doing that you should be learning how to develop your own business on the side in the field of study you love. That way you can just leave your corporate job and don't have to worry. There are a few red flags in your video about yourself. The fact that you are referring to your coworkers as family doesn't bode well. They are not your family, they are your coworkers. They have their own aspirations, which more than likely differ from your own. Your family are the kids you make and the man you marry. And considering you're 31 years old, the fact that you are calling your job 'family' tells me you have some misplaced maternal love.
With all due respect, I think you misunderstood. She's not saying that her co-workers are her family, she is saying "job family" as in "group of jobs".
On another note, your family is not necessarily "the man you marry and the kids you have", that is for each person to decide for themselves ;)
What we refuse to accept is the reality, that Data science, analysis and engineering were not meant to be standalone roles in the organisation. You are supposed to use Data related jobs as complementary skillsets like Excel, Word, PowerPoint and other programs to your other main job description. You can be a medical professional and data analyst with medical data. As an Educator, data science, analysis etc complement my student data processing and visualization. HR personnel needs the Data science/analysis skillset as a complement not stand alone. Else you will find yourself in and out of your primary job functions.
I'm 51yrs old and have been working in the operating room for 26yrs. Name a surgery and I've done it. Not a surgeon, but I work as their surgical assistant. Their right hand man, so to speak. I want to change careers and get out of the operating room. Can you explain how or where someone my age can begin a new path to being part of a data analyst with medical data like you mentioned?
@@mikel6787 There are lots of medical data that can be analysed in the field of medical surgery. You'll have to figure out how to gather relevant data in surgery that can be analyzed to improve surgical practices, save patients and make you the goto analyst consultant for a fee.
Heya! I dream to be a data scientist but i'm bit confused what degree program I should take.I mean is it better to go for 4 years software engineer degree then do masters in data science or do 4 year degree in data science and apply for data science job?Please help
Have you decided?
Yes, you work with a very trendy name, and few problems you mentioned, comes with trends!
You are too Good :)
I believe there is a massive discrepancy in "the perception of a true Data Scientist role" between the industry and job candidates. When you use data to solve a business problem, it is data science.
On a high level, DS can be categorized as analytics, ML + inference focus, and data engineering (this is What LinkedIn and Airbnb categorized). When you say "A true Data Scientist" in your above video, you are referencing "ML or casual inference focused" DS. However, the DS field has become more specific compared to maybe 10 years ago when only the ML-focused roles were called DS.
When you say you got asked only SQL (and I assume product case questions, like a consulting interview), you refer to a "Data analyst" role. However, these are also called Data scientists with an analytics focus. The boundary between data analyst and analytics DS becomes blurred, and I will use seniority and scope of work to differentiate them.
If you are solely searching for ML focused DS, you can apply for "Machine Learning Engineer" or DS with an emphasis on modeling in their job description.
I do have to mention that the (Product) Analytics-focused DS has a very large demand. Meta, LinkedIn, Airbnb, DoorDash, Uber, Spotify.. all these big names are hiring for these positions.
hello. I read your comments. what would you suggest for someone 29 years old lawyer (masters) who just wants to get into the field? would u suggest getting into data science and working for law firms? analyzing data to help the firm grow?
hello. I read your comments. what would you suggest for someone 29 years old lawyer (masters) who just wants to get into the field? would u suggest getting into data science and working for law firms? analyzing data to help the firm grow?
One more serious reason not to join DS (and to become Salesforce dev) is that managers enforce to find correlations in datasets where there is no correlation. In such a cases data scientists find out that "NO is also the answer" only in science, not in business. I saw one data scientists completely in depression because his manager insisted that after the advertising campaign there MUST BE a surge in profits, but there was no surge, and the manager insisted that this data scientist is just a bad professional. I also attended a lecture by the chief data scientist of Atlassian, where he told how for years he was bringing statistics to his managers that they did not like. In other words, data scientists often have to bring BAD NEWS to their superiors, and they REALLY don't like!
To be fair you said it yourself already that data science actually covers a family of domains or jobs. But I think indeed DS job is still too new and really rides on a lot of hypes and so a lot of the companies just didn't want to be left behind yet at the same time didn't really know what they're doing themselves. Also in reality there is yet to see that many companies (or even industries) truly benefit from having a DS or even just ML team from the business standpoint or profitability point.
Yaayyyy!! I am your biggest fan Sundas! love your informative videos as always!
All data scientists with RUclips channels explaining all carrer's pros and cons are doing god's work for us that are trying to break into the field. Thank you for your time and shared experiences.
I just left a DS role and took another job as a data engineer.
Lots of companies think they need DS, but aren't quite sure what to do with the DS they have on staff. Personally, I ended up doing a lot of ad hoc, proof-of-concept projects. But most of my workday involved assembling data in SQL for domain experts to review. When you're the only guy on the team who knows SQL, SQL is all you'll do :)
how's it possible to be the only guy on the team who knows SQL? I'm guessing the company does not have any software dev/engineers?
@James Page I was a DS on a team where everyone else was a domain expert.
@@MartyAckerman310 how does one become a domain expert without tech skills in the first place? Did they all have MBAs?
@@HarshvardhanKanthode Domain expert != coding and data skills.
Most of the people I was working with were engineers with deep product knowledge, but without coding and data skills.
fungible
I can’t with that word anymore 😂 gives me trauma 😭😂
As an outsider to the software/tech world, I am envious that discussing mental health, toxic work environments, mental happiness is even a thing. In every other industry (except TV news and mainstream news media) if you brought up those terms people would first look at you like you were an alien from Mars. Then everyone would start laughing at you for being such a soft woosie.
i think the data scientist -> product manager thing is a natural progression for people with quantitative skills who are also interested in entrepreneurship. they start out by wanting to apply math to business and then realize how much more interesting the business side of things really is.
I completely agree to what you say , the issue is the existing product manager are unable to think datascience solutions for the product. I have seen a lot of products where they should consider AI features primary they end up treating those as secondary. This makes their product outdated.
I understand what you said about applied scientist, but its a hard fact.
I hear ya. I saw a lot of job posting recently with title Software Engineer but job description is more on the data analyst side. I recently applied and interviewed for a Software Engineer position and turned out during the interview, the hiring manager told me he wasn't sure why programming languages are in the description, but his team only deal with SQL and data like 90%.
The whole applied science role seems pretty typical for tech, sounds like the DS version of full-stack. Get more done with less, treat people like resources...not people etc. Everything you said seems accurate and everyone has at least one bad experience in tech. I think everyone in tech should have a back up plan/exit plan when they are entering in to it. Either moving up to management or something different.
It's because managers continuously work with data. Data is a measurement of how a product or service performs so keeping an eye on data is very important.
Nice! I see you use the Momentum extension in chrome. I've used it for years. Such a good extension. Soo thank you for this video. You gave some good potential warnings. I am completing the Google Data Analytics cert and think it might help me in my current job. I planned on taking a Data Science cert after but I'm not sure. It depends on how things go but perhaps I still will. I'm confident I'm capable, even after listening to your points. But I think I have a better expectation now. Thank you :)
I preioct that applied scientists are just going to become the new data scientists. Most companies that aren't FAANG level expect you to have a somewhat broad field of expertise. If your team isn't huge, you're expected to take care of many different kinds of projects that come up.
It was like I was listening to my own story. My dumb manager with a software engineering background never understood data and data science. And then he totally pivoted the team's purpose where I was left underutilized. However, I got a good role after all this happened, but the transition time was unimaginably killing and stressful.
The role of data scientists in my company is quite ambiguous. My job as a data scientist is more like a tester. Some of my colleagues whose job title is also data scientist are more like developers. I spent most of my time studying business logic, looking for something inappropriate in the current system, and trying to find a better way to build analytic models. Some of my colleagues spend most of their time realizing the model, tuning parameters, and refining models. If we are going to turn the analysis into a product then I do most of the design as well as quality control work like a product manager and my colleagues do most of the implementation work. I guess it's more appropriate to give us different titles, like data analyst? scientist? engineer? I need to turn to a dictionary to tell the difference lol.
Could data science be interrupted by the constant growth of gpt's