LangGraph Comparative Review Agent

Overview

This repository contains a LangGraph agent that, given three entities, builds a comparative review table. The agent:

1. Generates 3‑5 comparison criteria using an LLM.
2. For each entity‑criterion pair performs a Tavily web search and collects a short note.
3. Builds a Markdown table with rows = criteria, columns = entities.
4. Produces a short verdict recommending the best entity for each use case.

The agent is implemented in src/main.py and can be run from the command line.

Requirements

pip install -r requirements.txt

The following environment variables must be set:

• OPENAI_API_KEY – key for the OpenAI model used by LangChain.
• TAVILY_API_KEY – key for the Tavily API.

Usage

# Default demo – compares Chroma, FAISS, Qdrant
python -m src.main

# Custom entities
python -m src.main "Entity1" "Entity2" "Entity3"

The script prints the generated criteria, the comparison table, and the verdict.