Stop building shit nobody wants. LaunchMintAI is a high-octane research engine that uses dual-layer search grounding and parallel agentic analysis to tear your ideas apart and rebuild them into viable business models.
- Validator: Real-time market data extraction (TAM/CAGR) using Tavily + Gemini 1.5 Flash. No hallucinations, just grounded data.
- War Room (Corporate Spy): Infiltrate the competition. Get deep-dive financials and "kill strategies" for incumbents.
- VC Roast (The Skeptic): A ruthless analysis of your fatal flaws. If you can survive the roast, you might survive the market.
- Pitch Forge (The Salesman): Instant, high-conversion taglines, elevator pitches, and value props.
- Frontend: Vite + React + Tailwind + Lucide (Premium UI/UX)
- Backend: FastAPI (Python) + Unified Extension System
- LLM: Google Gemini 1.5 Flash / 2.0 Flash
- Search: Tavily AI (God Mode grounded search)
git clone https://github.com/Jatin23K/LaunchMintAI.git
cd LaunchMintAIcd backend
python -m venv venv
source venv/bin/scripts/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txtCreate a .env in the backend/ folder:
GEMINI_API_KEY=your_gemini_key
TAVILY_API_KEY=your_tavily_keyRun the server:
python -m uvicorn app.main:app --reload --port 8000cd frontend
npm installCreate a .env in the frontend/ folder:
VITE_GEMINI_API_KEY=your_gemini_keyRun the app:
npm run devThe system uses a Waterfall Search Strategy:
- Tier 1 Authority: Scrutinizes McKinsey, BCG, Gartner, and Statista first.
- AI Judge: Every search result is semantically audited by a separate LLM pass to filter out SEO garbage.
- Math Fallback: If sources are missing Current TAM but have Forecasts + CAGR, the engine calculates the missing data to ensure a logical growth narrative.
LaunchMintAI provides strategic insights based on public data signals. It does not replace terminal-velocity execution or the founder's grit. Use it to build better, move faster, and fail less.