Modified End To End Resume ATS Tracking LLM Project With Google Gemini Pro

Поделиться
HTML-код
  • Опубликовано: 12 янв 2025

Комментарии • 130

  • @GAurAv-tn7wg
    @GAurAv-tn7wg 11 месяцев назад +3

    this project is very good sir .
    Thanks for solving poppler problem I am getting it since that video..
    I am ai and data science engineering student and I am in final year .
    Thanks for giving such a good knowledge abou genai❤

  • @DigitaTransforma
    @DigitaTransforma 11 месяцев назад +3

    This is a lot better. In your previous video, i was wondering why you had to convert the pdf to image.
    Extracting the text from the pdf was what I thought would be better.
    Thanks for the modification.

  • @argish
    @argish 11 месяцев назад +6

    Act like a skilled or very experienced ATS (Application Tracking System) with a deep understanding of tech field, software engineering, data science, data analyst and big data engineer. Your task is to evaluate the resume based on the given job description. You must consider the job market is very competitive and you should provide best assistance for improving the resumes. Assign the percentage Matching based on Jd and the missing keywords with high accuracy.
    resume: {text}
    description: {jd}
    I want the response in one single string having the structure \
    {{"JD Match":"%",
    "Missing Keywords : []",
    "Profile Summary" : ""}}
    Thank me later.

    • @durgeshjaiswar3321
      @durgeshjaiswar3321 11 месяцев назад +1

      what is Google API, and where to get API?

    • @arshasangral5785
      @arshasangral5785 11 месяцев назад

      @argish hii I was trying to understand the code and it’s really fascinating to see how generativeAI works. I just have one question from where Ai model is taking text and jd values when we have not given any parameters to AI model. Thankyou

    • @argish
      @argish 11 месяцев назад

      @@arshasangral5785 if you take a look at the website UI, there's a text box. We have to write jd in there. Since it is similar to us chatting with Gemini, only parameters we can pass are text (in our case, jd) and the resume pdf. The text we write in the text box will be added to the prompt template that we built (specifically, it'll replace jd in ' description = {jd} '). Finally, the whole prompt will be passed on to the model along with the pdf.

  • @amitguitarist2008
    @amitguitarist2008 10 месяцев назад

    Thanks Krish . It is always great to follow your tutorials. I think we need to format the prompt before passing to the model, for eg. input_prompt_complete = input_prompt.format(extracted_text=extracted_text,job_description=job_description) ...

    • @NaturesCall5
      @NaturesCall5 5 месяцев назад

      @amitguitarist2008 why its giving different output for same js and resume

  • @nadipudisriharshini4649
    @nadipudisriharshini4649 9 месяцев назад +1

    Hiii Krish I have done what u have said in this video i got the output but the output doesn't match correctly for Resume and Job description if there is no match in the resume and jd also it is showing 75%. Can you please tell us

  • @nagbhushanrsubbapurmath2247
    @nagbhushanrsubbapurmath2247 11 месяцев назад

    Yes Now I got it, Thank you for your Video in this!!!!!.....Wanted to see more videos on the Gen AI Like this!!!😇

  • @inetselection
    @inetselection 8 месяцев назад

    Thank you and small correction as underlined
    def input_pdf_text(uploaded_file):
    reader=pdf.PdfReader(uploaded_file)
    print(len(reader.pages))
    print(reader.pages[0])
    text=""
    for page in (len(reader.pages)):
    page=reader.pages[page]
    text+=str(page.extract_text()) if page.extract_text() else ""
    return text

  • @mikew2883
    @mikew2883 11 месяцев назад +1

    This is great! I ended up setting up the code but for some reason the analysis is wildly different each time you scan the same resume. For instance the percentage goes from a range of 60-74% and the missing keywords are non-existent in the original job description.

    • @durgeshjaiswar3321
      @durgeshjaiswar3321 11 месяцев назад

      what is Google API, and where do we find it?

    • @francis6303
      @francis6303 11 месяцев назад

      That job simply doesn’t suite for you 😜

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      Hey did you get any blank screen when you run the app.py file?

  • @vineethaburugupalli2635
    @vineethaburugupalli2635 11 месяцев назад +5

    @Krish Naik. Many of us getting error ###AxiosError: Request failed with status code 403. Could you please help to resolve the error

    • @AnandChavan1999
      @AnandChavan1999 10 месяцев назад +1

      Have you solve your problem .. I am facing same issues

    • @mohamed_deshaune
      @mohamed_deshaune 10 месяцев назад

      ​@@AnandChavan1999hey this problem is from streamlit itself you can solve it by following this :
      first create a folder and name it .streamlit
      then create a file inside the folder and name it config.toml
      and put this line inside the file above:
      [server]
      enableXsrfProtection = false
      enableCORS = false

    • @MB-sh9ur
      @MB-sh9ur 9 месяцев назад

      me too

    • @govindjha1323
      @govindjha1323 6 месяцев назад

      do you get the solution of it??

  • @arpitkhare4286
    @arpitkhare4286 8 месяцев назад +2

    model displaying totally random range from 55 to 85 for the same JD and cv.

  • @RavitejaGundimeda
    @RavitejaGundimeda 11 месяцев назад +2

    Hey krish, it is showing the keywords that are not present in job description as missing! Also the percentage was varying when I did the submission twice! Just saying!

  • @AyushKumar-fx8bn
    @AyushKumar-fx8bn 11 месяцев назад

    If facing blank screen after running streamlit then make sure you mentioned proper model name. It should be '('gemini-pro')'. In video spelling mistake present

  • @UrCloud2
    @UrCloud2 11 месяцев назад

    Love your projects. So informative. Thanks

  • @satyamedicherla3236
    @satyamedicherla3236 9 месяцев назад +1

    Hi Krish, why am I getting this error "AttributeError: module 'google.generativeai' has no attribute 'GenerativeModel'"? Any changes/updates in the API?

  • @ramisettysathwik2834
    @ramisettysathwik2834 11 месяцев назад +3

    Hey Krish Can you do a video on extensions which you are using which will be quite useful for us and the way of explaining from the scratch is the best i've ever seen. Thank you

  • @rachrach9871
    @rachrach9871 5 месяцев назад

    I encountered the same errors but I installed poopler in my venv using
    conda install -c conda-forge poppler and everything worked perfectly

  • @rajeshkannan8608
    @rajeshkannan8608 11 месяцев назад +3

    AxiosError: Request failed with status code 403
    This error come whenever i try to submit my resume

    • @AnandChavan1999
      @AnandChavan1999 10 месяцев назад

      Hello You have got solution related same error

    • @AnandChavan1999
      @AnandChavan1999 10 месяцев назад

      Please reply

    • @mohamed_deshaune
      @mohamed_deshaune 10 месяцев назад

      ​@@AnandChavan1999the problem is from streamlit itself you can solve it by :
      first create a folder and name it .streamlit then create a file inside it and name it config.toml and put this line inside it :
      [server]
      enableXsrfProtection = false
      enableCORS = false

    • @darshitshah8668
      @darshitshah8668 8 месяцев назад

      Even I am facing same error, any solution for this?

    • @govindjha1323
      @govindjha1323 6 месяцев назад

      @@darshitshah8668 have u got the solution

  • @REFIREGAMING
    @REFIREGAMING 11 месяцев назад +1

    There are a couple of problems with this code, this generates a random response
    here are some of the errors - there is a corrected version of this code on my github if anyone is interested.
    our text field is not being used by the program since there is no option for gemini to access out input prompt from the text field
    same for the resume that is uploaded
    so as a result, it generated random responses

    • @syuvaraj4791
      @syuvaraj4791 11 месяцев назад

      Hey. I suppose you have run the program. Can you let me your LinkedIn or something? I have a few doubts about running it.

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      Hey did you get any blank screen when you run the app.py file?

    • @chinmayoutube
      @chinmayoutube 11 месяцев назад

      I do wish to get the corrected version of this code. Can you share the same? Github link or LinkedIn, anything will do.

  • @rakeshkumar-pf6yu
    @rakeshkumar-pf6yu 7 месяцев назад

    Can you please make a video of RAG model using Multiple PDF files , and all files having data as (Text + Tables + Images ) in all the PDF files. This is what they are asking in interview these days almost in every interview. Also they ask about the different LLM app Evaluation techniques and in which case which technique to use....Please make a detailed video for these two topics...Thanks...I m a BIG FAN of yours Sir....

  • @yashasravichandra8445
    @yashasravichandra8445 11 месяцев назад +1

    Hey Krish, I just want to inform you that here in London we don't have access to create a google api key, I know it is not possible to use GPT api keys for a gemini ,please give us a solution regarding this issue or can you please make this project using GPT

  • @rockstar3191
    @rockstar3191 11 месяцев назад +5

    anybody getting error while uploading pdf? like status code 403? can you say whats the issue?

    • @namangupta1249
      @namangupta1249 9 месяцев назад

      yes i am getting - AxiosError: Request failed with status code 403

    • @govindjha1323
      @govindjha1323 6 месяцев назад

      @@namangupta1249 did you got the answer?

  • @mansikamthane3788
    @mansikamthane3788 2 месяца назад

    Hiii krish. Every time I submit the app with same JD and same resume , it provides different results. Sometimes present keys in resume too shows as missing keys

  • @adityachakole755
    @adityachakole755 10 месяцев назад

    Hey Krish, thank you for this !!

  • @pratapsimha1466
    @pratapsimha1466 4 месяца назад

    Hi krish, why don't you create a docker imagr of this and push it to docker hub?

  • @decodingdatascience
    @decodingdatascience 11 месяцев назад +1

    Issues with pypdf2, python 3.10 @krish

  • @koustavbanerjee3824
    @koustavbanerjee3824 10 месяцев назад

    Hey Krish.Can you upload a video on how we can extract text from scanned pdf documents and do the analysis like bank statement which is in a tabular format.Thank you

  • @lixiasong3459
    @lixiasong3459 11 месяцев назад

    This is amazing, thank you, Sir!

  • @AIandSunil
    @AIandSunil 11 месяцев назад +1

    Update 1. The get_gemini_response function is using input as a parameter, but it's not being passed to the function. You should pass the input_prompt as an argument when calling the function.
    Update 2. In the line st.subheader(response), it seems like you want to display the response as a subheader, but you forgot to format it with the input parameters.

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      Hey did you get any blank screen when you run the app.py file?

    • @rajeshkannan8608
      @rajeshkannan8608 11 месяцев назад

      AxiosError: Request failed with status code 403
      This error come whenever i try to submit my resume

  • @pareeksachin_23
    @pareeksachin_23 11 месяцев назад +2

    Hello Sir,
    After I run this app, give my resume and jd, and receive the results, the results are always changed like missing words and percentages.
    Why did it happen.

    • @nilesh30300
      @nilesh30300 11 месяцев назад +2

      I'm also getting a similar issue. For the same JD and same resume, at first, it shows a 60% match, and next time it shows 86%. Also, it shows some missing keywords which are already present in the resume. Results vary a lot so could you please let us know what could be the reason. Thanks

    • @Sourav_Samanta.
      @Sourav_Samanta. 11 месяцев назад

      ​@@nilesh30300even I also face same problem 😅I uploaded love babbar SDE sheet of DSA ... it's show 54 percentage match lol 😂

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      When i run the app.py file, it is showing the blank screen, there was no error displayed. Please can you help me with these

    • @pareeksachin_23
      @pareeksachin_23 11 месяцев назад

      @@ShashiCodes Bro just follow sir steps and re-run your code

  • @madmusician67
    @madmusician67 8 месяцев назад

    hello sir i have built this project and it is working fine but there are some bugs in the project like it is showing the output even when the uploaded pdf doesn't contain a resume

  • @2faced_001
    @2faced_001 10 месяцев назад

    this line of code: """ for page in reader(len(reader.pages)): """
    I getting this error: 'PdfReader' object is not callable
    'PyPDF2._reader.PdfReader' object is not supposed to accept this integer argument 'len(reader.pages)' and turn into an iterable.
    Kinda confused somebody help! ;__;
    (update> fixed my issue): If I'm not wrong it's supposed to be **for page in range(0,len(reader.pages)): **

  • @DevtuneAI
    @DevtuneAI 6 месяцев назад

    @krish
    I getting below error, even I tried a lot, in python-3.10
    AttributeError: module 'google.generativeai' has no attribute 'GenerativeModel'

  • @AkshanshDhama
    @AkshanshDhama 11 месяцев назад

    Why new courses are not being added to Tech Neuron now nor a single live class ?

  • @dakshpatel4557
    @dakshpatel4557 6 месяцев назад +1

    can we do this using langchian ? sir please explain this and build the same application using langchainn

  • @techn0Spikes
    @techn0Spikes 11 месяцев назад

    Thanks and appreciate this. However getting AxiosError: Request failed with status code 403 while uploading file.. Any solutions suggested?

  • @NoDoglapan
    @NoDoglapan 11 месяцев назад

    Hi Krish - make one video for a streamlit app that takes some params and invokes a model (pkl) file. thanks

  • @mraabhijit
    @mraabhijit 11 месяцев назад +1

    Great resource Krish. I am getting a weird error while running the app. It returns a halucinated text as resume content and in no way matches the resume I upload to it. Any specific reason for this?

    • @rohithakash8093
      @rohithakash8093 7 месяцев назад

      That's because the input such as resume isn't going with the prompt, Go ahead and pass it to the function of gemini along with your prompt

  • @LokeshKumar-ds4gb
    @LokeshKumar-ds4gb 11 месяцев назад

    Somehow, it is giving a irrelevant response, I am applying for a devops engineer roles, and mentioned the same in the script. Though my JD doesnt have 10+ year of experience because i am applying for a associate/mid level, It still gives this output in profile summary:
    "The candidate possesses 10+ years of experience in software development and DevOps practices. Demonstrated expertise in managing CI/CD pipelines, implementing continuous deployment strategies, and ensuring seamless software delivery. Proven ability to collaborate with cross-functional teams to drive successful software development projects. Seeking a fast-paced environment to leverage strong technical skills and enhance team productivity."}}

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      Hey did you get any blank screen when you run the app.py file?

    • @LokeshKumar-ds4gb
      @LokeshKumar-ds4gb 11 месяцев назад

      @@ShashiCodes no it didn't give any blank

  • @anna-archieve
    @anna-archieve 11 месяцев назад

    what if everyone uses genAI to write resumes!!!! then what ATS will do? every resume for a particular domain will have same probability to get selected!!, so what companies will do then?

  • @kanishkasharma353
    @kanishkasharma353 11 месяцев назад

    sir I have been following with the genAI tutorials for a long time but still none of the apps are working
    kindly make a vdo on how to fix errors.

  • @vishyt4884
    @vishyt4884 11 месяцев назад +1

    Hi sir i am getting 400 api key not valid at the end can anyone suggest me to solve the error

  • @crank_tmf
    @crank_tmf 11 месяцев назад

    Unable to get page count. Is poppler installed and in PATH? getting this error at ats resume expert what to do

  • @AmrutaGurav-z6z
    @AmrutaGurav-z6z 10 месяцев назад

    Hi krish i have tried this and whenever i add same JD it gives different ans like JD mtach sometime 65% sometime 85% and same with keywords also. why like this??

  • @tasneemshaikh5139
    @tasneemshaikh5139 11 месяцев назад

    Can u please tell how to run any GitHub downloaded project in our systems what are all the directories we have to make please sir

  • @chiranjeevikumarbattula8709
    @chiranjeevikumarbattula8709 11 месяцев назад +1

    Sir again submitted same resume and job description it varies the results that me once it give 75% and again run it gives 65% . 😢
    What we need to do in this situation

    • @krishnaik06
      @krishnaik06  11 месяцев назад

      Work on the prompt template

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      @@krishnaik06 When i run the app.py file, it is showing the blank screen, there was no error displayed. Please can you help me with these.

  • @NitinSingh-vi1xi
    @NitinSingh-vi1xi 11 месяцев назад

    how to use it? cuz i havent started any programming learning yet any alternative for me?

  • @prasaddugyala7330
    @prasaddugyala7330 11 месяцев назад

    Sir,iam completed Mcom in 2019.i have 4 years career gap for govt jobs preparation.i want to learn AI,ML. And data analytics.which is best for me?
    How much time required to get a job?

  • @kunalaggarwal2574
    @kunalaggarwal2574 11 месяцев назад

    Sir I am continuously getting python fatal error: Modulenotfounderror: no module named "io".

  • @ujjwalrajput7809
    @ujjwalrajput7809 11 месяцев назад

    I'm unable to get the api as it shows you're not authorize to this

  • @aswinkumardamarapalli9585
    @aswinkumardamarapalli9585 11 месяцев назад

    Hi @krish.. i have one doubt. While using LLM (gpt or gemini) we have one paramater temperature. While developing one application like a chatbot i tuned this parameter, when i set temp low i got a response like, 'hi, i am ai assistant how can i help you.' . When i set temp high ' hi, i am brsin tumor assistant which i can help you with reports, scans etc.......'. how can temperature affecting the result. Can you brief this parameter and any other parameters while tuning LLM. thanks in advance

    • @AimTMR
      @AimTMR 11 месяцев назад +1

      Temperature introduces randomness in answer. You can say more temperature more creative answers.

    • @saikiranedits
      @saikiranedits 11 месяцев назад

      ​@@AimTMR yes 💯 percent correct

    • @aswinkumardamarapalli9585
      @aswinkumardamarapalli9585 11 месяцев назад

      @@AimTMR I understand that . I am asking what's the behind logic for generating that. Infact more temperature sometimes leads to irrevalant answers as I practically checked it. If we understand logic on how itz working, we will do better. That's the thought

    • @durgeshjaiswar3321
      @durgeshjaiswar3321 11 месяцев назад

      what is Google API, and where to get API?

  • @ShashiCodes
    @ShashiCodes 11 месяцев назад

    Hey Krish, I did everything shown here but when i run app.py file ,the screen was blank. What might be the reason?

  • @yashwantsahu4069
    @yashwantsahu4069 11 месяцев назад

    Thanks Sir, It's Working.

  • @sanjay.r2584
    @sanjay.r2584 11 месяцев назад +1

    Do anyone know y i get this error..? ###AxiosError: Request failed with status code 403

    • @rajeshkannan8608
      @rajeshkannan8608 11 месяцев назад

      yes i get same error

    • @govindjha1323
      @govindjha1323 6 месяцев назад

      @@rajeshkannan8608 did you got the solution??

  • @phaneendrag6021
    @phaneendrag6021 11 месяцев назад

    I tried to run on Google Colab, it threw up several errors. Can anyone please help me out with it.

  • @tmo185
    @tmo185 11 месяцев назад

    hello krish naik i have one query is that if you will put only two word for example (python,sql) in job description then also it will give this output:-{{"JD Match":"80%","MissingKeywords":["Software Development Life Cycle","Agile Methodology","Cloud Computing","Machine Learning Algorithms","Data Mining","Tableau"],"Profile Summary":"Seasoned professional with 10+ years of experience in software engineering and data science. Proven expertise in developing and deploying complex software systems, working with large-scale data sets, and applying machine learning techniques to solve business problems. Demonstrated ability to lead and mentor teams, collaborate effectively with cross-functional teams, and deliver high-quality results within tight deadlines."}} ,why it is providing irrelevant feedback which is not related to job description

    • @aniBunny12378
      @aniBunny12378 10 месяцев назад

      Gemini is telling you what is missing from the prompt and not the JD.

  • @hitthegold
    @hitthegold 11 месяцев назад

    Thank you for this video. I am now able to run my own ATS app on my system. Great help from you, Krish! I checked your earlier videos and found these missing parameters in this video code:
    def get_gemini_repsonse(input_prompt,):
    model=genai.GenerativeModel('gemini-pro')
    response=model.generate_content([input_prompt,])
    return response.text

    • @ShashiCodes
      @ShashiCodes 11 месяцев назад

      Hey did you get any blank screen when you run the app.py file?

  • @TarotbyTanyaa
    @TarotbyTanyaa 8 месяцев назад

    Hi guys , please help to troubleshoot this error"DefaultCredentialsError: "

  • @durgeshjaiswar3321
    @durgeshjaiswar3321 11 месяцев назад

    what is Google API, and where to get API?

  • @jaskarankukreja8852
    @jaskarankukreja8852 10 месяцев назад

    Is this api key available in canada?

  • @sanjaykhadka5952
    @sanjaykhadka5952 11 месяцев назад

    @@krishnaik06 can we read excel file also?

  • @User_11185
    @User_11185 11 месяцев назад

    how to run this app. I have no idea how to do that.

  • @GungunSangwan-q9s
    @GungunSangwan-q9s 2 месяца назад

    Please help me every time is okay but i cant understand the last Step

    • @GungunSangwan-q9s
      @GungunSangwan-q9s 2 месяца назад

      My program run in terminal itself
      no new tab open

  • @ARkhan-xw8ud
    @ARkhan-xw8ud 11 месяцев назад

    Can you make a custom chatbot using gemini api

  • @mohammadanas7929
    @mohammadanas7929 11 месяцев назад

    thank you sir

  • @umerulla
    @umerulla 11 месяцев назад +1

    I did everything shown here but when i run app.py file the screen is blank.. what may be the issues i am facing?

  • @shighrasahil2504
    @shighrasahil2504 6 месяцев назад

    Can you please create this same system using a free llm with huggingface

  • @udaykiran6474
    @udaykiran6474 11 месяцев назад

    If i ask question in hindi spanish Gemini pro doesnt work

  • @ShobhaKumariChoudhary-m2b
    @ShobhaKumariChoudhary-m2b 11 месяцев назад +1

    Sir can u please guide on how to run this app on our laptop. tried to run this code but not able to execute it.

  • @Guitartoons
    @Guitartoons 11 месяцев назад +1

    Since you are typing by looking at an existing code... It's better you just explain the code. Since I guess you ain't well versed with the code . Just a suggestion

  • @KumR
    @KumR 11 месяцев назад

    i think this belongs to gemini playlist

  • @explorewithskp1237
    @explorewithskp1237 11 месяцев назад

    Sir it is giving wrong output. I am giving data scientist job description which has no big data related info but in output is mentioning them
    {{"JD Match":"70%","MissingKeywords":["Distributed Systems","AI/ML","Cloud Computing"],"Profile Summary":" Experienced software engineer with a strong background in data analysis and big data engineering. Proven ability to develop and implement efficient data processing solutions using Hadoop, Spark, and other open-source tools. Seeking to leverage my skills in a challenging role where I can make a significant contribution to a team."}}

  • @rassbrothers
    @rassbrothers 11 месяцев назад

    I just join ineuron for only krish sir

  • @Doomedragon
    @Doomedragon 10 месяцев назад

    Comes up with irrelevant response all the time

  • @tusharvarun-f6q
    @tusharvarun-f6q 10 месяцев назад

    hey krish im still getting AxiosError: Request failed with status code 403

  • @chandinisaikumar2736
    @chandinisaikumar2736 9 месяцев назад

    Hey @krishnaik06, While I am trying to upload my resume, I am getting an error
    AxiosError: Request failed with status code 403
    Can you please help me out in this

    • @MB-sh9ur
      @MB-sh9ur 9 месяцев назад

      can you tell how you run this code?

  • @roblars9252
    @roblars9252 11 месяцев назад

    and here I extended on the application 🤓 even added a few extra prompts installed poppler as a package to streamlit and added a webhook functionality with automation even added a personalized touch with each prompt wanting the user to feel welcomed, I only get api errors but not a lot I figured it was the amount of api calls I made but you might have better insights to it I dont want to leave it like that.. What is your email? I can send you mine I tried separating the files too so I can extend on the application I want to build a real big time ATS System

    • @olanrewajuatanda533
      @olanrewajuatanda533 10 месяцев назад

      Hi, can I send you my email for your edited files?

  • @nagrajkaranth123
    @nagrajkaranth123 11 месяцев назад

    Thank you sir