| name | langchain-retrieval-agent |
| description | AI agent with retrieval tool for document Q&A using RAG and LangGraph. |
LangChain Retrieval Agent
An AI agent with retrieval tool for document Q&A using RAG, powered by LangGraph.
Tech Stack
- Framework: Next.js
- AI: LangChain.js, LangGraph, 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-agent.git .
If the directory is not empty:
git clone --depth 1 https://github.com/Eng0AI/langchain-retrieval-agent.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
SUPABASE_URL- Supabase project URLSUPABASE_PRIVATE_KEY- Supabase service role keyOPENAI_API_KEY- For embeddings and LLM
Build
pnpm build
Development
pnpm dev