Media innovation in minutes
summary
A MapReduce style platform called Archer runs user-supplied Docker images and ffmpeg recipes at serverless scale so hundreds of thousands of CPUs finish prototyping in minutes. The system splits media, maps user code per split, and collects outputs while using fuse-based mounting to lazy-download frames. Production numbers include 13 billion frames processed, 2 billion text detections, 770 million dialogue analyses and 120 million shot-change events, with compute metering and access controls.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:57Use cases
- 1:11Artwork
- 2:46Episode Art
- 3:22Subtitle Authoring
- 4:02Archer
- 5:40Netflix
- 6:16Grayscale
- 9:18Platform
- 15:59Content security
related talks
30mTaking Ruby to the Movies
43mConference Recording and Streaming
25mContent creation under a new creative paradigm from on-premises to cloud
44mRelease After a Year of Development
36mBuilding a Self-Serve Stream Processing Platform with Flink
09mWhy I ditched cloud transcoding and set up a mac mini render farm
22mBuilding a Flexible, Scalable Infrastructure to Test Dolby.IO at Scale
24mLive Video Transmuxing/Transcoding