LangChain Tutorial (TypeScript): building a RAG pipeline for intermediate developers

This tutorial builds a production-shaped Retrieval-Augmented Generation pipeline in TypeScript using LangChain: load documents, split them, embed them into a vector store, and answer user questions with retrieved context. You need this when a plain LLM is not enough and you want answers grounded in your own PDFs, docs, or knowledge base instead of model memory.

What You'll Need

• •Node.js 18+ installed
• •A TypeScript project initialized
• •An OpenAI API key exported as OPENAI_API_KEY
• •Packages:
  • •langchain
  • •@langchain/openai
  • •@langchain/community
  • •@langchain/core
  • •typescript
  • •tsx or ts-node for running TypeScript directly
• •A source document to index, for example:
  • •PDF
  • •.txt
  • •Markdown files

Install the dependencies:

npm install langchain @langchain/openai @langchain/community @langchain/core
npm install -D typescript tsx @types/node

Step-by-Step

1. •Start by loading a document from disk and splitting it into chunks. RAG works better when you retrieve smaller, focused chunks instead of stuffing whole documents into the prompt.

import { TextLoader } from "@langchain/community/document_loaders/fs/text";
import { RecursiveCharacterTextSplitter } from "langchain/text_splitter";

async function main() {
  const loader = new TextLoader("./data/policy.txt");
  const docs = await loader.load();

  const splitter = new RecursiveCharacterTextSplitter({
    chunkSize: 800,
    chunkOverlap: 120,
  });

  const chunks = await splitter.splitDocuments(docs);
  console.log(`Loaded ${docs.length} doc(s), split into ${chunks.length} chunk(s)`);
}

main();

2. •Next, embed the chunks and store them in a vector database. For local development, an in-memory store is fine; in production, swap this for Pinecone, pgvector, or another persistent backend.

import { TextLoader } from "@langchain/community/document_loaders/fs/text";
import { RecursiveCharacterTextSplitter } from "langchain/text_splitter";
import { OpenAIEmbeddings } from "@langchain/openai";
import { MemoryVectorStore } from "langchain/vectorstores/memory";

async function buildIndex() {
  const loader = new TextLoader("./data/policy.txt");
  const docs = await loader.load();

  const splitter = new RecursiveCharacterTextSplitter({
    chunkSize: 800,
    chunkOverlap: 120,
  });

  const chunks = await splitter.splitDocuments(docs);
  const embeddings = new OpenAIEmbeddings({ model: "text-embedding-3-small" });
  const vectorStore = await MemoryVectorStore.fromDocuments(chunks, embeddings);

  return vectorStore;
}

buildIndex().then(() => console.log("Vector index ready"));

3. •Now wire retrieval to an LLM. The retriever finds the most relevant chunks for a question, and the model uses those chunks as context to generate an answer.

import { ChatOpenAI } from "@langchain/openai";
import { createStuffDocumentsChain } from "langchain/chains/combine_documents";
import { createRetrievalChain } from "langchain/chains/retrieval";
import { ChatPromptTemplate } from "@langchain/core/prompts";
import { TextLoader } from "@langchain/community/document_loaders/fs/text";
import { RecursiveCharacterTextSplitter } from "langchain/text_splitter";
import { OpenAIEmbeddings } from "@langchain/openai";
import { MemoryVectorStore } from "langchain/vectorstores/memory";

async function main() {
  const loader = new TextLoader("./data/policy.txt");
  const docs = await loader.load();

  const splitter = new RecursiveCharacterTextSplitter({ chunkSize: 800, chunkOverlap: 120 });
  const chunks = await splitter.splitDocuments(docs);

  const embeddings = new OpenAIEmbeddings({ model: "text-embedding-3-small" });
  const vectorStore = await MemoryVectorStore.fromDocuments(chunks, embeddings);
  const retriever = vectorStore.asRetriever(4);

  const llm = new ChatOpenAI({ model: "gpt-4o-mini", temperature: 0 });

  const prompt = ChatPromptTemplate.fromMessages([
    ["system", "Answer only using the provided context. If the context is insufficient, say you don't know."],
    ["human", "{input}"],
    ["system", "Context:\n{context}"],
  ]);

  const combineDocsChain = await createStuffDocumentsChain({ llm, prompt });
}

4. •Finish by invoking the retrieval chain with a real question. This is the part you will actually call from your app or API route.

import { ChatOpenAI } from "@langchain/openai";
import { createStuffDocumentsChain } from "langchain/chains/combine_documents";
import { createRetrievalChain } from "langchain/chains/retrieval";
import { ChatPromptTemplate } from "@langchain/core/prompts";
import { TextLoader } from "@langchain/community/document_loaders/fs/text";
import { RecursiveCharacterTextSplitter } from "langchain/text_splitter";
import { OpenAIEmbeddings } from "@langchain/openai";
import { MemoryVectorStore } from "langchain/vectorstores/memory";

async function main() {
  const docs = await new TextLoader("./data/policy.txt").load();
  

Keep learning

• •The complete AI Agents Roadmap — my full 8-step breakdown
• •Free: The AI Agent Starter Kit — PDF checklist + starter code
• •Work with me — I build AI for banks and insurance companies

By Cyprian Aarons, AI Consultant at Topiax.