Langchain State Management with React App

[suryaiyer95]

I have created a few agents and want to host them as a REST service with a React frontend. Few questions

Q1 However, I'm unsure how to handle state management when multiple users are using the agents together. Since the agents have memory and a vector store, how can I ensure that each user's interaction with the agents is isolated and doesn't impact the experience of other users?

Q2: When hosting my agents as a REST service with a React frontend, how can I implement a retry mechanism to resume where the agents left off in scenarios where the user refreshes the frontend? what would be the best way to handle the state of the agent? which would be the memory, goals, result uptill now. and the any changes in the vector store

Are there any best practices or open source projects that could help with this? Thanks and appreciate your help and suggestions!

[rafaelquintanilha]

Haven't tried anything similar, but from what you described, my first try would probably set up a Redis Memory and use the user id to identify each particular history.

This should address both your questions, once the LLM would store each interaction in the appropriate key and this could be recovered at any time.