EPG API — Electronic Program Guide

The Project

We developed a high-performance REST API that serves as the central data source for electronic program guides (EPG). The system supplies multiple media platforms simultaneously with TV schedule data, broadcast details, recommendations, and streaming information — reliably, scalably, and with response times under 125 milliseconds.

Challenge

TV schedule data is complex: hundreds of thousands of broadcasts with over 130 attributes per record, linked with genres, ratings, cast lists, and external databases. The API must prepare this data for different clients — each with their own field mappings, slugs, and feature sets — while delivering stable response times even under load. At the same time, recommendations, search features, and accessibility features such as audio descriptions need to be seamlessly integrated.

Core Features

• EPG Grid API: Optimized endpoints for program grid displays — delivers broadcasts by channel groups, time windows, and dates with configurable entry counts and intelligent time range logic.
• Multi-Tenant Architecture: The same system serves multiple clients with individual configurations, slug systems, and feature sets — without code duplication.
• Personalized Recommendations: Daily top recommendations and category-based program suggestions, curated by broadcast type and highlight rating.
• Streaming Integration: Proxy connection to external streaming services to display on which platforms content is available — with a dedicated caching layer.
• Full-Text Search: Efficient search across broadcasts, series profiles, and programs with filtering by date, channel, and category.
• Dual API Versioning: Parallel V1 and V2 APIs with full backward compatibility — enabling gradual client migration without breaking changes.
• Accessibility: Integration of audio description metadata and specialized broadcast types for an inclusive user experience.

Technical Highlights

• 75-90% faster queries through 16+ optimized composite indexes on the critical query paths of the broadcast table
• Multi-tier Redis caching strategy with content-dependent TTLs — 24 hours for program grids, 12 hours for profiles, 1 week for master data
• Automatic response compression via gzip middleware with 60-80% bandwidth reduction
• 60+ API endpoints with complete OpenAPI documentation, rich examples, and validation rules
• Service layer architecture with clean separation of HTTP logic, business rules, caching, and validation
• Multi-database connectivity with parallel access to primary and secondary databases for optimized read operations
• Response times under 125ms for the most-used grid listing endpoints in production

Result

The API is the backbone of multiple media platforms, delivering TV schedule data to millions of end users daily. Through the combination of intelligent indexing, multi-tier caching, and response compression, the system achieves stable sub-125ms response times — even under heavy load. Dual API versioning enables evolutionary development without disruption to existing clients.