Bitu Technology is the registered company of Tubi's China team. We are engineers passionate about leveraging advanced technology to improve lives, and we hope to use this channel to connect and advance together.
The Tubi Meetup explored how AI coding tools are reshaping software engineering, emphasizing a shift from prompt engineering to long‑term validation, error‑feedback loops, and team organization, and questioning how small teams can sustainably serve millions of users in the AI era.
This recap of Chen Tian's AI‑focused meetup explores why knowledge still matters, presents four practical learning methods, outlines effective AI prompting techniques, redefines the developer’s role from coder to conductor, and offers a five‑stage AI collaboration workflow for modern software creation.
This article describes how Tubi improved the latency of its Redis‑backed online feature store for machine‑learning inference by analyzing query patterns, measuring client‑side bottlenecks, and applying optimizations such as binary Avro encoding, MGET usage, virtual partitioning, and parallel deserialization to meet a sub‑10 ms SLA.
The article recounts a Tubi meetup where senior engineers presented Argos, a cloud‑native load‑testing platform built with curl, Lambda, and ClickHouse to handle Super Bowl‑level traffic, while also discussing engineering mindset, cross‑team processes, and how AI tools empower developers to become 10x engineers.
The article details how Tubi's Web/OTT team restructured player error reporting, statistical metrics, and unified handling, introduced precise error‑tracking enums, defined new recovery strategies for device decoding, network, and cache issues, and validated their impact through extensive experiments that improved user experience and key business KPIs.
This article describes how Tubi's Android team redesigned their video player with a modular, distributed architecture—introducing PlayerContext, PlayerHandlerWrapper, and composable modules for ads, event tracking, and TTS—to improve flexibility, testability, and scalability while sharing lessons learned from the refactoring process.
The Tubi Android team replaced a dual‑player architecture with a single Media3/ExoPlayer instance, achieving smoother video‑ad transitions, reduced buffering, lower error rates, and improved business metrics, while also addressing decoder‑initialization failures, non‑key‑frame ad buffering, and subtitle timing issues.
This article explains how Tubi monitors the quality of its recommendation system by identifying potential failure points, tracking key data streams such as model input, final recommendation output, and training data, and designing a scalable, real‑time monitoring solution with clear protocols and extensible metrics.
This article describes how Tubi built the Rosetta Stone system—a flexible ID mapping workflow that leverages large language models, embedding similarity ranking, and K‑nearest‑neighbors to unify and enrich metadata across a 200,000‑title library, improve content recommendation, and streamline operations.
The article recaps the ninth Scala Meetup where speaker Pei Qi (Li Guobin) introduced GraphQL‑driven partial POJOs, Scala 3's Mirror feature, Magnolia as an alternative, type‑class concepts, and compared Scala 2 and Scala 3, while also promoting the Tubi data‑platform lead role and past meetup resources.
