Activity
Mon
Wed
Fri
Sun
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
What is this?
Less
More

Memberships

AI Developer Accelerator

10.8k members • Free

2 contributions to AI Developer Accelerator
ANY IDEAS: CrewAI picking API endpoints
Hey guys I need some help figuring out the best way to get an agent or task to pick out the best API endpoint to call based off the users input. There will be around 14 endpoints and each one contains information in json form. There is a LOT of data in each endpoint and so i want the agent to return the endpoint containing the info the user wants i.e the email of "John" etc. What is the best way to give the agent or task the knowledge/ framework of each endpoint so that it will accurately give the correct endpoint back to me everytime?
0 likes • Feb 3
@Bastian Venegas would appreciate if we could pm or jump in a call as i would really benefit with some more specific help!!
0 likes • Feb 3
@Bastian Venegas Cant dm you sadly just realised not a high enough level, ill explain it here, as mentioned i have around 14 endpoints, let me describe a couple: One is for devices so its a json response that cointains the names, serial numbers the registration date. there is over 10000! Another one is for licenses, containing the license name, status (in compliance or out of compliance) and quantity of licenses. My aim would be to ask a question like: What licenses do I have that are out of compliance? Ive begun learning about flows and just cant wrap my head around how I would actually pass the info so the llm can comprehend it, I'm assuming you were thinking about using a router in the flow but how would that actually work? And then additionally i would create a seperate tool to call each endpoint and get the response?
Chatbot agent exceed token limit
Hello everyone! I’m working on the LangChain Master Class for Beginners 2024 video, and I’m struggling with rag_conversational. Instead of using the provided text files, I input a large document (The Capital, Volume I—around 500 pages) and attempted to request a summary of Chapter 1 using OpenAI. However, I encountered an error message. Does anyone have any ideas on how to resolve this issue? message': 'Request too large for gpt-4o in organization org-aVsRZmS23mqZY3EFUoQNNDCw on tokens per min (TPM): Limit 30000, Requested 542017. The input or output tokens must be reduced in order to run successfully. Visit https://platform.openai.com/account/rate-limits to learn more.', 'type': 'tokens', 'param': None, 'code': 'rate_limit_exceeded'
0 likes • Feb 3
Could always try running a local LLM as opposed to using openai, something like llama3.2 using ollama!! Or write a function to break the document down and parse it in!
1-2 of 2
Jean Mady
1
3points to level up
@jean-mady-4239
Uni Student who is an aspiring software engineer :)

Active 314d ago
Joined Jan 23, 2025
Powered by