Back to selected work
04 // Backend engineering · Media discovery
Braindance API
A media API for discovery, franchise collections, device-scoped watchlists and playback-progress telemetry.
Context
Streaming clients need fast catalogue discovery, curated franchise grouping and personalized watchlists/resume-playback state without forcing full user account onboarding.
Contribution
- Built a modular Express + TypeScript backend integrating movie and television metadata from the TMDB API.
- Implemented search, discovery, franchise collections, anonymous device-scoped watchlists and playback-progress telemetry endpoints.
- Added API-key middleware, rate limiting, Zod request validation, Pino structured logging and SWC-powered builds.
Key engineering decisions
- Cached frequently queried genre metadata in MongoDB with TTL-based expiry to reduce redundant upstream TMDB calls.
- Scoped watchlists and playback progress to anonymous device identifiers so client apps can persist user state without mandatory auth.
- Versioned API routes and standardized Pino structured logs so route evolution and operational debugging stay isolated.
Outcomes
- Provided a clean, versioned media discovery and telemetry service ready for client streaming interfaces.