"I couldn't explain it to the freshmen. That means we don't really understand it." — Richard Feynman

Non-Tech Speak is an AI-powered platform designed to bridge the gap between technical complexity and simple understanding. Inspired by Richard Feynman's philosophy, it transforms dense technical jargon into relatable, context-aware analogies for various audiences—from a 5-year-old child to a busy CEO.

• Context-Aware Translations: Converts technical concepts into analogies tailored to specific personas (e.g., "Like a pizza delivery service" for APIs).
• Multi-Persona Support: Explanations for different audiences: 5-Year-Old, Grandmother, Non-Tech Manager, etc.
• Real-time Streaming: Instant feedback using AI streaming (Google Gemini 2.0 Flash).
• Social Learning: Public feed of popular translations, upvoting system, and shared wisdom.
• Atomic Design UI: Beautiful, consistent interface built with Shadcn UI and Tailwind CSS.
• Performance First: Built on Bun and ElysiaJS for maximum speed and low latency.

We prioritized speed, type safety, and modern developer experience.

Category	Technology	Description
Runtime		Ultra-fast JavaScript runtime & bundler.
Frontend		App Router for reactive, server-first UI.
Backend		High-performance TypeScript framework.
Database		Serverless Postgres managed via Drizzle ORM.
AI Engine		Powered by Google Gemini via @tanstack/ai.
Auth		Robust, type-safe authentication.
Styling		Utility-first CSS with Shadcn UI.
Rate Limiting		Redis for rate limiting and caching.

The project follows a Monorepo-style hybrid architecture, separating the Next.js frontend from the Elysia backend logic while keeping them in the same codebase for cohesion.

.
├── app/                  # Next.js App Router (Frontend)
│   ├── (auth)/           # Authentication routes
│   ├── (dashboard)/      # Protected dashboard & app core
│   ├── api/              # Next.js API routes (proxies)
│   └── page.tsx          # Landing page
├── server/               # Content & Backend Logic (Elysia)
│   ├── routes/           # API Endpoints (Translation, History, etc.)
│   ├── db/               # Database Schema & Drizzle Config
│   └── services/         # Business logic & AI Integration
├── components/           # Atomic Design System
│   ├── atoms/            # Basic building blocks (Buttons, Inputs)
│   ├── molecules/        # Combinations (Search bars, Cards)
│   ├── organisms/        # Complex sections (Forms, Sidebars)
│   └── templates/        # Page layouts
├── public/               # Static Assets
└── lib/                  # Shared Utilities

Follow these steps to get the project running locally.

• Bun: This project uses Bun as the package manager and runtime.
• PostgreSQL: Ensure you have a Postgres instance running (or use Neon).

1. Clone the repository

   git clone https://github.com/yourusername/non-tech-speak.git
   cd non-tech-speak

2. Install dependencies

   bun install

3. Set up Environment Variables Copy the example environment file and fill in your secrets.

   cp .env.example .env.local

   See .env.example for the full list of required keys (Database, AI Keys, Auth Secrets).

4. Initialize the Database

   bun run db:generate
   bun run db:push

5. Run the Development Server

   bun dev

   Open http://localhost:3000 to see the app.

• Hybrid Architecture: We use Next.js for SEO and initial rendering, while heavy logic and AI streaming are handled by ElysiaJS. This gives us the best of both worlds: great SEO and raw API performance.
• Feynman Technique: The core of the app is built around the idea that "complexity is a mask for lack of understanding." The AI is prompted not just to translate, but to teach.

Contributions are welcome! Please run bunx biome check . before submitting a PR to ensure code quality.

This project is licensed under the MIT License.