Build Amazing Multi Agent Crews and Monitor them - Latest CrewAI and LangWatch Integration
ฝัง
- เผยแพร่เมื่อ 29 ก.ย. 2024
- In this video, we’re thrilled to introduce two fantastic new additions to Langflow: CrewAI and LangWatch! These tools are set to make your experience with multi-agent systems and retrieval-augmented generation (RAG) even better.
Key learnings:
- How to build CrewAI agent crews completely no code in Langflow
- How to build Sequential Agent Crews that takes a prompt and completes tasks in sequence
- How to build hierarchical Agent Crews like Managers and Workers
- How to auto-generate prompts for Agents
- How to connect LangWatch with Langflow and start logging, tracing, debugging your LLM / RAG Apps
- How to create evals for your LLM / RAG apps using LangWatch
... and much more
This is the only video that shows a tutorial with CrewaI on Langflow, and it is terrible. We also have a Sequential Task Agent.
Is this example available to play with? Where can I find it thanks,, don't see the Langwatch component in Langflow ?
so why it is not working for me?
none of the tempaltes works
Langflow rocks!!! It's beautiful, powerful and simple!!!! Amazing!!!!
Thanks!
Hi! My Lanfglow is version 1.0.9 and I don't have there the 3 examples on the bottom: "Sequential Tasks Agent", "Hierarchical Tasks Agent" and "Complex Agent".
It is just a demonstration or this is on a newer version? Witch version?
Great questions
Sentiu falta também, né?
Rs@@AlexBrumMachadoPLUS
it's available from version 1.0.10
@@Langflow great, tks
How do you compare with flowise and langgraph? Thanks you so much
Amazing Video. Please create more in-depth tutorials around CrewAI agents
In version 1.012, how do I attach an image/photo and ask Gemini 1.5 flash or GPT-4.0 to analyze? It was working before the update; at least the LLM could receive and read what I sent.
Oh no! That is not good! Hopefully it gets fixed man.
How to route request in Langflow based on the "intent"? What I mean is to have AI evaluate input, and recognize intent. for example intent might be: Support, Finance, Sales etc. Then based on recognized intent it routes to different part of the flow. Voiceflow has quite good and simple to use implementation of intent if need for reference. How to achieve something similar with Langflow?
Looking for the same too for my chatbot