Project a

This repository provides a small full‑stack example using a TypeScript Express backend and a React + Vite frontend. The directories are organised by responsibility:

src/
  bot/
    core/
    flows/
    nlp/
  ui/
    components/
    themes/
  api/
public/
examples/
.github/
  workflows/

The backend exposes several endpoints under /api:

• /api/joke returns a random joke.
• /api/weather returns current weather information.
• /api/foia generates a FOIA request letter when provided request details.

The React application located in my-app/ includes a simple chat UI that calls these endpoints.

Additional capabilities include:

• Voice input and output powered by the Web Speech API.
• Basic multi-language support via a small i18n dictionary.
• Embedding-based search with cosine similarity.

Development

Install dependencies and start the development server:

npm install
npm run dev

In a separate terminal build and serve the React frontend:

cd my-app
pnpm install
pnpm dev

The Vite development server proxies requests starting with /api to http://localhost:3000, so make sure the backend is running with npm run dev in another terminal.

Run npm run build to compile the backend to dist/.

Examples

A simple p5.js demo is available in examples/p5-spiral. Open index.html in a web browser to see the animated spiral.

License

This project is released under the MIT License.

RL Scheduler PoC

A lightweight Python 3.11 micro-service lives under src/green_ai/. It exposes a FastAPI app with two endpoints:

• POST /step – performs one reinforcement-learning step using rack telemetry from Prometheus.
• GET /metrics – Prometheus-scrapable metrics for Grafana dashboards.

Running locally

poetry install
uvicorn src.green_ai.service:app --reload

Synthetic load simulation:

make simulate