AI Startup Investment Copilot 🚀

An Intelligent Deal Sourcing & Analysis Platform for Venture Capitalists

This project is a high-fidelity demo designed for a VC daily work. it demonstrates the power of Multi-Agent Systems (MAS) in processing unstructured data, conducting deep industry research, and performing quantitative investment evaluations.

🌟 Hero Features

• Agentic Deal Sourcing & Research: Automatically retrieves web data via Serper and transforms cluttered web pages into clean, LLM-ready Markdown using Firecrawl.
• 3-Agent Collaborative Workflow (LangGraph):
  • 🕵️ Researcher Agent: Conducts broad searches and deep web crawling.
  • 📊 Analyst Agent: Extracts key metrics (TAM/SAM/SOM, Team Strength, Product Moat, Traction).
  • ⚖️ Investment Committee Agent: Performs rigorous fact-checking and scoring based on a custom VC Rubric.
• Investment Dashboard: A minimalist, high-performance dashboard built with Next.js + Shadcn/UI, featuring real-time AI report rendering.
• Data Persistence: Hybrid storage of structured profiles and unstructured analysis results using Supabase (PostgreSQL).

🚀 Getting Started

Prerequisites

• Python 3.10+
• Supabase CLI (for local database development)
• Node.js & npm/pnpm (for the upcoming frontend)

1. Backend Setup

Navigate to the backend directory and set up your environment:

cd src/backend

# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Configure environment variables
cp .env.example .env
# Edit .env with your API keys (OpenAI, Serper, Firecrawl, Supabase)

2. Database Setup (Supabase)

This project uses Supabase for data persistence. You can run it locally or use a remote project.

Local Development:

# Initialize Supabase (if not already done)
supabase start

# Apply migrations
supabase db reset

Manual Schema Setup: If you are using a remote Supabase project, run the SQL found in supabase/migrations/20260305071128_init_schema.sql in your Supabase SQL Editor.

3. Running the Application

Start Backend API

cd src/backend
uvicorn main:app --reload

The API will be available at http://localhost:8000. You can access the interactive documentation at http://localhost:8000/docs.

---

🏗️ Architecture Overview

Tech Stack

• Frontend: Next.js 14/15 (React), Tailwind CSS, Shadcn/UI, Lucide-React.
• Backend: Python 3.10+, FastAPI, LangGraph (Agentic Orchestration), LangChain.
• AI Stack: GPT-4o / Claude 3.5 Sonnet, Firecrawl (Web Scraping), Serper (Search API).
• Database: Supabase (PostgreSQL + JSONB for flexible reporting).
• Deployment: Vercel (Frontend), Railway/Docker (Backend).

LangGraph Workflow

1. Input: Startup Name + Website URL + Industry.
2. Step 1 (Researcher): Serper (5-10 links) -> Firecrawl (Clean Markdown).
3. Step 2 (Analyst): Analyze Markdown -> Extract Market, Team, Product, and News signals.
4. Step 3 (Investment Committee): Fact-checking -> Rubric-based Scoring -> Risk Flagging.
5. Output: Multi-tab Investment Memo + Final Deal Score (e.g., 7.8/10).