Build a Simple AI RAG System with LangChain 1.0

Level: Real World

Your task for today is to create a minimal document chat assistant using LangChain v1 that loads your documents and answers questions about them using core LangChain concepts.

Project Task

Create a minimal document chat assistant that:

• Loads all PDF, text (.txt), and markdown (.md) files from a folder

• Combines all documents into a single context string

• Uses LangChain’s init_chat_model to initialize the AI model

• Sends the entire context to an AI model with system messages

• Answers questions about your documents in a conversational way

• Maintains message history for context

Basic Features:

• Load multiple file types (PDFs, text files, markdown files)

• Simple document loading - no complex processing or vector stores

• Uses LangChain’s unified chat model interface

• Chat interface - ask questions naturally

• Message history - remembers previous questions and answers

• Works with OpenAI or Google GenAI models

• Easy to use - just load and chat!

This project introduces you to core LangChain concepts, document processing, AI chat applications, and building practical AI tools — essential skills for working with LangChain and documents.

Expected Output

Chat with your documents! (type ‘quit’ to exit)

You: What are the land and building asset amounts of rabit in 2023

Assistant: Based on the documents, the land and building asset amounts for rabit in 2023 are:
- Land assets: $2.5 million
- Building assets: $5.8 million
Total real estate assets: $8.3 million

You: What was the total revenue?

Assistant: According to the documents, the total revenue for 2023 was $15.2 million.

You: quit
Goodbye!

Join the Python & AI Builders Skool Community

Got questions to ask the author about this project? Join our Python & AI Builders community for weekly Python & AI videos and discussions:

skool.com/automateit

💻 View This Project in Jupyter Notebook

Become a premium member to access the working code in a Jupyter Notebook with the ready-to-run solution to get the output instantly and extend the code as you wish.

Click below to access the notebook (7-day risk-free trial):