Problem
Deliver trusted, near-real-time meta insights and player profiles at scale while Riot’s APIs, patches, and traffic spikes evolve every set — all without sacrificing performance or reliability.
Teamfight.lol is a live TFT analytics platform tracking millions of players and matches per set. I’ve built the product end-to-end — from frontend and backend to data pipelines, infrastructure, and design. The platform delivers instant stats and fresh meta insights that the TFT community (from casual players to top-ranked pros) uses daily. It now serves as a trusted reference point across streams, Discords, and social media, proving both its scale and reliability.
Deliver trusted, near-real-time meta insights and player profiles at scale while Riot’s APIs, patches, and traffic spikes evolve every set — all without sacrificing performance or reliability.
Redis-first ingestion and query layer with RedisSearch; rate-limit-aware Riot API orchestration via Bull; Nuxt frontend tuned for instant search and big data views; custom admin for live ops; and automated static-data parsing/translation/normalization (AI-assisted).
The platform largely runs itself: continuous ingestion, resilient infra, and fast UI. Players get fresh stats in seconds; the team enjoys low day-to-day ops and rapid feature iteration.