Show HN: A Performance-Focused Watch Database Built with Next.js and Python

Hi HN,

I’ve been experimenting with building a lightweight, high-performance platform for exploring detailed watch specifications — focused on design structure, data modeling, and frontend rendering performance.

The project is live here: https://www.tagheuerreplica.io/

While it looks like a watch catalog, under the hood it’s a technical prototype designed to test: - *Next.js 14* for static generation and incremental revalidation - *FastAPI (Python)* as a microservice backend for structured metadata - *SQLite* for ultra-fast, read-optimized data queries - *Edge caching* via Vercel to maintain sub-100ms global load times

The architecture was built to explore how a small-scale app can serve complex reference data without heavy infrastructure or expensive cloud databases. It’s entirely serverless, and each watch reference loads from a precomputed JSON structure that syncs automatically when metadata updates occur.

I’m curious about feedback on: - How to further optimize SSR performance for large JSON datasets - Whether lightweight alternatives to SQLite (like DuckDB or Turso) might help - Edge function cold-start minimization strategies

This was a fun weekend project that grew into a benchmarking sandbox for watch-related metadata. Would love to hear any suggestions from the community — especially from those working on serverless data delivery or content-heavy apps.

Live demo: https://www.tagheuerreplica.io/