skypiea

Skypiea AI is a cloud‑native reference application that shows how to design, test, ship, and operate a Go service at production scale. Feel free to clone, reuse and feedback.

It also wraps the Gemini API to expose a built‑in chatbot interface. You can test it, free tier though...

Built with ❤️ in San Francisco

Production URL

Live demo: https://chat.skypiea-ai.xyz
Hosted on an EC2 t2.micro, secured by Cloudflare, served through Traefik with default certificates.

Routes

• / React landing (v2)
• /v1 HTMX UI
• /v2 React app
• /api/v2 JSON API

Why does this project exist?

Skypiea AI is not a traditional SaaS product that solves a single business problem. Instead, it is a living laboratory for modern backend engineering practices:

• Scalable – containerised, stateless by default, ready to autoscale in Kubernetes.
• Testable – unit, integration tests.
• Maintainable – layered architecture, and uniform logging, managed database migrations on startup, run on local easily with little dependency.
• Secure – AuthN/Z implementation of JWT and session and similar sensitivity on prod; HTTPS and secrets stored encrypted on k8s/DB etc...
• Automated – full CI/CD pipeline that lints → tests → builds → pushes → deploys on every git push.
• Portable – runs the same on your laptop or in the cloud.

Overall, think of Skypiea AI as a live resume: a solid demonstration of senior-level backend practices.

Tech Stack

Quick Start (Local)

# 1. Clone
$ git clone https://github.com/fukaraca/skypiea && cd skypiea

# 2. Spin up a DB and make migration
$ make migratedb-up

# 3. build images and run the server
$ make dc-build-up

# 4. visit http://localhost:8081 in your browser (React UI); v1 is at /v1

V2 UI Dev

# 1. Install UI dependencies
$ make ui-install

# 2. Run the Vite dev server (http://localhost:5173)
$ make ui-dev

The dev server proxies /api/v2, /v1, and /healthz to http://localhost:8080.

License

Skypiea AI is licensed under the MIT License. See LICENSE for details.

Contact

Maintainer: Linkedin