Traction Backend

A platform that gives founders a web presence for their ideas before they build anything - shareable, discoverable, and credible. By allowing founders to easily structure their ideas, receive AI-generated feedback, and instantly produce an elegant web summary and pitch deck, Nebular Blazar replaces scattered PDFs and slides with a single, modern URL.

Features

• Ideation Assistant: Chat with an AI that helps refine messy ideas into structured business models while offering honest readiness flags.
• Dynamic Deck Generation: AI automatically generates rich, interactive, full-screen HTML/CSS/JS pitch decks based on structured data.
• Investor Summary: Instantly highlights strengths, concerns, and key metrics in a simple view.
• llm.txt Endpoints: Provides a machine-readable startup spec, optimized for AI agents discovering projects.
• Effortless Sharing: Share one link that includes both the generated pitch deck and summary.

Tech Stack

• Backend: FastAPI (Python)
• Database: PostgreSQL
• ORM: SQLModel / SQLAlchemy
• Database Migrations: Alembic
• AI Models: OpenAI API, Pydantic-AI
• Authentication: JWT, Google OAuth 2.0 Authlib
• Templating: Jinja2 (server-side rendered views)
• Environment Management: uv (Python package manager)

Prerequisites

• Python 3.12+
• uv package manager
• Docker and Docker Compose (for running PostgreSQL locally)

Getting Started

1. Clone the repository

   git clone <your-repo-url>
   cd traction-backend

2. Configure Environment Variables Create or modify the .env file in the root directory to ensure all necessary keys are populated. Typical required values include database credentials, secret keys, Google API credentials, and your OpenAI API key.

3. Start the Database Using Docker Compose, you can quickly spin up a local PostgreSQL instance:

   docker-compose up -d

4. Install Dependencies Install the Python environment and dependencies using uv:

   uv sync

5. Start the Development Server Start the FastAPI development server with Uvicorn:

   cd backend
   uv run uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

   The backend will be available at http://localhost:8000.

Project Structure

• backend/app/: Core FastAPI application logic (controllers, models, API routes, templates, etc.)
• alembic/ (if configured): Database migrations
• docker-compose.yml: Local database orchestration
• pyproject.toml & uv.lock: Dependency definitions and locks

---

Note: This backend also handles the frontend rendering using server-side Jinja2 templates for public views. Everything operates securely within the FastAPI structure without requiring a separate frontend deployment.