Convex Chat Memory

For longer-term persistence across chat sessions, you can swap out the default in-memory chatHistory that backs chat memory classes like BufferMemory for Convex.

Setup

Create project

Get a working Convex project set up, for example by using:

npm create convex@latest

Add database accessors

Add query and mutation helpers to convex/langchain/db.ts:

convex/langchain/db.ts

export * from "langchain/util/convex";

Configure your schema

Set up your schema (for indexing):

convex/schema.ts

import { defineSchema, defineTable } from "convex/server";
import { v } from "convex/values";

export default defineSchema({
  messages: defineTable({
    sessionId: v.string(),
    message: v.object({
      type: v.string(),
      data: v.object({
        content: v.string(),
        role: v.optional(v.string()),
        name: v.optional(v.string()),
        additional_kwargs: v.optional(v.any()),
      }),
    }),
  }).index("bySessionId", ["sessionId"]),
});

Usage

Each chat history session stored in Convex must have a unique session id.

tip

See this section for general instructions on installing integration packages.

npm

npm install @langchain/openai @langchain/community

Yarn

yarn add @langchain/openai @langchain/community

pnpm

pnpm add @langchain/openai @langchain/community

convex/myActions.ts

"use node";

import { v } from "convex/values";
import { BufferMemory } from "langchain/memory";
import { ChatOpenAI } from "@langchain/openai";
import { ConversationChain } from "langchain/chains";
import { ConvexChatMessageHistory } from "@langchain/community/stores/message/convex";
import { action } from "./_generated/server.js";

export const ask = action({
  args: { sessionId: v.string() },
  handler: async (ctx, args) => {
    // pass in a sessionId string
    const { sessionId } = args;

    const memory = new BufferMemory({
      chatHistory: new ConvexChatMessageHistory({ sessionId, ctx }),
    });

    const model = new ChatOpenAI({
      model: "gpt-3.5-turbo",
      temperature: 0,
    });

    const chain = new ConversationChain({ llm: model, memory });

    const res1 = await chain.invoke({ input: "Hi! I'm Jim." });
    console.log({ res1 });
    /*
      {
        res1: {
          text: "Hello Jim! It's nice to meet you. My name is AI. How may I assist you today?"
        }
      }
    */

    const res2 = await chain.invoke({
      input: "What did I just say my name was?",
    });
    console.log({ res2 });

    /*
      {
        res2: {
          text: "You said your name was Jim."
        }
      }
    */

    // See the chat history in the Convex database
    console.log(await memory.chatHistory.getMessages());

    // clear chat history
    await memory.chatHistory.clear();
  },
});

API Reference:

• BufferMemory from langchain/memory
• ChatOpenAI from @langchain/openai
• ConversationChain from langchain/chains
• ConvexChatMessageHistory from @langchain/community/stores/message/convex