Connect Flowise to Google Sheets using webhooks and monitor with LangSmith
HTML-код
- Опубликовано: 4 сен 2023
- In this video we connect Flowise to Google Sheets using make.com webhooks. We Connect Flowise to LangSmith to monitor the communication and observe the arguments passed in function callings.
01:11 Create the data structure as a table in Google Sheets
03:41 Use make.com template with webhook and Google Sheets
04:14 Use Postman to send a post tothe make.com webhook
06:27 Setup Flowise with LangSmith
13:12 Check Tokens used and Function Calling Args in LangSmith
Another great video. I really appreciate the diverse use-case examples that introduce the tools that expand the programing potential, especially to us newbs that don't know about their existence or function. You are single-handedly expanding by toolkit. Thanks again.
Glad it was helpful! There are so many possibilities when you use webhooks to connect Flowise to other services.
Great video! Thanks!
Glad you liked it!
excellent, thanks brotha!
Thank you for your feedback.
Great video, Flowise added the image functionality, how can I capture images in the google sheet? Or any other format?
Thanks. Have not checked it yet.
Hi. New to flowise. How does the 'trigger' happen in flowise ? Can we use a method in the parameters. eg look for 'save in notes' ?
Hi, yes. Please check function calling.
Exactly what I was looking for. Could you do one with Notion, too? Thank you!
Glad you liked it. A Video or Short about connecting to Notion is in planning.
@@business24_ai plssss, I need it!
As a realtor, could i make this send webhooks with specific info like new home buyers? Trying to think how to parse the info so the bot can have a casual convo but capture the needed variables and transfer to google sheets
Hi, your question has two part. One part is the trained model which extract the useful real estate information, and the second part is the transfer of this information to a service like Google sheet. This video covers the second part. For the first part you can go to hugging face and search for a suitable solution for the real estate field.
Great video but you do understand that in order to get the LangChain API key you need to be subscribbed to their BETA, menaing it's not possible for those seeing now that don't have acees to complete the tutorial.. my case btw.
So ill wait to be accept and then give a more complete feedback.
Anyway thanks for the help and i'm looking foward to see more about Flowise
Thanks for your feedback. Unfortunately, LangSmith is not available publicly, so some parts of this tutorial may be helpful in the future. We hope this will change soon and you can complete the tutorial. Looking forward to your feedback once you have access.
This is pretty good. Thank you
If I wanted to log automatically.
Will adding a line "Always save the answer as a research note" trigger the custom function?
100% ensuring the function call is where I have struggled.
Glad you liked it. Please keep in mind that using custom functions is costly if you do use models that are not open-source. So if you need to log all of your answers I would suggest using another approach or using Python code directly and connecting to SQLite. Flowise supports databases like SQLite too. docs.flowiseai.com/databases
Thanks appreciate the response
Hi, regarding exporting all prompts and API messages please check ruclips.net/video/R-6uV1Cb8I8/видео.html
Good video and details but i fail to get the goal. I tried to write details of an appointment to google sheets with 5 columns tried to build.. Webhook works and Postman works as well in writing data to google sheets.. but Flowise couldn't save it. Just trying to save the Name, Age, Department, Remarks, Date and Time of Day. Not sure where its going wrong. Any specific prompts needs to be made so that it saves?
First of all, congratulations. Taking action and putting the knowledge to work itself is a goal and a milestone you have reached. The next Goal is to know if you need to connect your Flowise Flow to another app, how to do it, or where to look. Maybe, in your next project, you decide to put the appointment directly in a calendar app.
As for the logic of your AI App or bot, it's difficult to give an answer when there is no detail information. Improving the prompt can help, but if you need full control, you need to code in Python or JS.
Keep in mind to use a model that supports function calling. Models with suffix -0613 etc.
how do i save the whole question as well as the chatbot answer on google sheet every time a user asks a question? this seems to only work to process certain keywords in the question asked
I would use JavaScript and a database connection to SQLite to log both questions and answers and would not go over OpenAI to save tokens. Flowise supports Database connections docs.flowiseai.com/databases
Hi, to see and export all user prompts and the API answers, please check our new Video ruclips.net/video/R-6uV1Cb8I8/видео.html
I did everything you did but make is not receiving data from flowise. Any idea?
Hi, we moved our comments and discussion to our community on www.skool.com/business24ai/about. If you like, join the community while it is still free and ask the question over there.
I think you missed some steps in the postman part
Thanks for your feedback. At the end, you need to send a raw JSON post request to your webhook address with the structure. Feel free to use postman.com or insomnia.rest/ to send this post request. I plan to do another video with webhooks. I will consider this feedback and will do a more detailed step-by-step walkthrough to send the post request.
@@business24_ai i referred to the fact you used the lightweight client from the desktop app
I would like to consult with you regarding the tech part for my business. how can I contact you ?
Please use the E-Mail Address "info@business24.ai" to get in touch.
Hi, regarding exporting all user prompts and the API messages, please check ruclips.net/video/R-6uV1Cb8I8/видео.html