Skip to main content

Aurora DSQL Chat Memory

For longer-term persistence across chat sessions, you can swap out the default in-memory chatHistory for the serverless PostgreSQL-compatible Amazon Aurora DSQL Database.

This is very similar to the PostgreSQL integration with a few differences to make it compatible with DSQL:

  1. The id column in PostgreSQL is SERIAL auto-incrementent, and DSQL is UUID using the database function gen_random_uuid.
  2. A created_at column is created to track the order and history of the messages.
  3. The message column in PostgreSQL is JSONB, and DSQL is TEXT with Javascript parsing handling

Setup​

Go to you AWS Console and create an Aurora DSQL Cluster, https://console.aws.amazon.com/dsql/clusters

tip

See this section for general instructions on installing integration packages.

npm install @langchain/openai @langchain/community @langchain/core pg @aws-sdk/dsql-signer

Usage​

Each chat history session is stored in a Aurora DSQL (Postgres-compatible) database and requires a session id.

The connection to Aurora DSQL is handled through a PostgreSQL pool. You can either pass an instance of a pool via the pool parameter or pass a pool config via the poolConfig parameter. See pg-node docs on pools for more information. A provided pool takes precedence, thus if both a pool instance and a pool config are passed, only the pool will be used.

For options on how to do the authentication and authorization for DSQL please check https://docs.aws.amazon.com/aurora-dsql/latest/userguide/authentication-authorization.html.

The following example uses the AWS-SDK to generate an authentication token that is passed to the pool configuration:

import pg from "pg";

import { DsqlSigner } from "@aws-sdk/dsql-signer";
import { AuroraDsqlChatMessageHistory } from "@langchain/community/stores/message/aurora_dsql";
import { ChatOpenAI } from "@langchain/openai";
import { RunnableWithMessageHistory } from "@langchain/core/runnables";

import {
  ChatPromptTemplate,
  MessagesPlaceholder,
} from "@langchain/core/prompts";
import { StringOutputParser } from "@langchain/core/output_parsers";

async function getPostgresqlPool() {
  const signer = new DsqlSigner({
    hostname: process.env.DSQL_ENDPOINT!,
  });

  const token = await signer.getDbConnectAdminAuthToken();

  if (!token) throw new Error("Auth token error for DSQL");

  const poolConfig: pg.PoolConfig = {
    host: process.env.DSQL_ENDPOINT,
    port: 5432,
    user: "admin",
    password: token,
    ssl: true,
    database: "postgres",
  };

  const pool = new pg.Pool(poolConfig);
  return pool;
}

const pool = await getPostgresqlPool();

const model = new ChatOpenAI({
  model: "gpt-4o-mini",
});

const prompt = ChatPromptTemplate.fromMessages([
  [
    "system",
    "You are a helpful assistant. Answer all questions to the best of your ability.",
  ],
  new MessagesPlaceholder("chat_history"),
  ["human", "{input}"],
]);

const chain = prompt.pipe(model).pipe(new StringOutputParser());

const chainWithHistory = new RunnableWithMessageHistory({
  runnable: chain,
  inputMessagesKey: "input",
  historyMessagesKey: "chat_history",
  getMessageHistory: async (sessionId) => {
    const chatHistory = new AuroraDsqlChatMessageHistory({
      sessionId,
      pool,
      // Can also pass `poolConfig` to initialize the pool internally,
      // but easier to call `.end()` at the end later.
    });
    return chatHistory;
  },
});

const res1 = await chainWithHistory.invoke(
  {
    input: "Hi! I'm MJDeligan.",
  },
  { configurable: { sessionId: "langchain-test-session" } }
);
console.log(res1);
/*
  "Hello MJDeligan! It's nice to meet you. My name is AI. How may I assist you today?"
*/

const res2 = await chainWithHistory.invoke(
  { input: "What did I just say my name was?" },
  { configurable: { sessionId: "langchain-test-session" } }
);
console.log(res2);

/*
  "You said your name was MJDeligan."
*/

// If you provided a pool config you should close the created pool when you are done
await pool.end();

API Reference:

Was this page helpful?


You can also leave detailed feedback on GitHub.