AI Startup Validator

Overview

AI Startup Validator is a multi-agent system that analyzes startup ideas from multiple perspectives—market research, financial planning, tech architecture, and risk assessment—using advanced LLMs (via LangChain, Ollama) and real-time web search. The system provides actionable insights and recommendations to help entrepreneurs make informed decisions about their startup concepts.

---

Features

• Market Research Agent: Identifies competitors, trends, and market gaps using live web search.
• Financial Planning Agent: Estimates costs, suggests optimizations, and analyzes funding risks.
• Tech Architect Agent: Designs scalable tech stacks and infrastructure recommendations.
• Risk Assessment Agent: Evaluates business, technical, financial, and market risks.
• CEO Agent: Synthesizes all reports to make a final decision and strategic roadmap.
• Feedback Loop: If high risk is detected, the system iterates for improved recommendations.

---

Installation Guide

1. Clone the Repository

git clone <your-repo-url>
cd ai-validator

2. Set Up Python Virtual Environment

python -m venv .venv
.venv\Scripts\activate  # On Windows
# or
source .venv/bin/activate  # On Mac/Linux

3. Install Requirements

pip install -r requirement.txt

4. Install and Run Ollama (for LLM backend)

• Download and install from ollama.com
• Start Ollama and pull the required model:

ollama
ollama run llama3.1

5. Run the App

streamlit run app.py

---

Usage

1. Enter your startup idea in the text area.
2. Click Run Multi Agent Analysis.
3. View detailed reports from each agent and the CEO's final recommendation.
4. All results are saved in shared_memory.json for future reference.

---

Project Structure

ai-validator/
├── agents.py              # Multi-agent system logic
├── app.py                 # Streamlit web app
├── tools.py               # Utility functions (web search, cost calculation)
├── requirement.txt        # Python dependencies
├── shared_memory.json     # Stores last analysis results

---

Requirements

• Python 3.8+
• Ollama (for local LLM inference)
• Internet connection (for web search)

---

Example Startup Idea

A platform that connects local farmers directly with consumers, providing fresh produce and reducing food waste.

---

References

• LangChain Docs
• Ollama
• Streamlit
• DuckDuckGo Search

---