LangChain Retrieval

Document Q&A with RAG (Retrieval Augmented Generation) using Supabase vector store.

Tech Stack

• Framework: Next.js
• AI: LangChain.js, AI SDK
• Vector Store: Supabase pgvector
• Package Manager: pnpm

Prerequisites

• Supabase project with pgvector extension
• OpenAI API key

Setup

1. Clone the Template

git clone --depth 1 https://github.com/Eng0AI/langchain-retrieval.git .

If the directory is not empty:

git clone --depth 1 https://github.com/Eng0AI/langchain-retrieval.git _temp_template
mv _temp_template/* _temp_template/.* . 2>/dev/null || true
rm -rf _temp_template

2. Remove Git History (Optional)

rm -rf .git
git init

3. Install Dependencies

pnpm install

4. Setup Environment Variables

Create .env with required variables:

• SUPABASE_URL - Supabase project URL
• SUPABASEPRIVATEKEY - Supabase service role key
• OPENAIAPIKEY - For embeddings and LLM
• SUPABASEDBURL - Direct PostgreSQL connection URL

5. Setup Vector Store

Initialize pgvector extension and create documents table in Supabase.

Build

pnpm build

Development

pnpm dev