For our upcoming talk at KubeCon AI day co-located event, we plan to demo a ReAct agent setup that can be easily reproduced by others for building an agent based LLM application. To implement the demo we should:
[x] Define a use case we want to showcase
[x] Implement a ReAct agent using LangChain
[x] Define a set of tools for the agent to interact with
[x] Deploy MlFlow to log the agent's outputs and decision making traces
hemajv
commented
2 weeks ago
For our upcoming talk at KubeCon AI day co-located event, we plan to demo a ReAct agent setup that can be easily reproduced by others for building an agent based LLM application. To implement the demo we should: