Serverless at the end of the Universe
summary
Claim: serverless struggles with state so code ships data to remote stores, hurting latency and complexity. Presents research fixes like Shredder and Cloudburst, and practical tooling: Azure Durable Functions for orchestrations, Custom Vision AI for model training. Builds a Spirit Dog app that trains models (parallel tag uploads, then train) and a Durable Days scheduler, debugs locally with Core Tools and VS Code, and demonstrates orchestration, timers and human‑interaction workflows that actually run end‑to‑end.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:10Presentation
- 10:12Shredder
- 26:15Orchestrators
- 27:53Train Model
- 29:17Conclusion
- 30:14Questions
- 38:27Outro
related talks
28mThe Brave New World of serverless computing
40mThe State of Serverless Computing
51mConfusion In The Land Of The Serverless
58mServerless - the brief past, the bewildering present, and the beautiful (?) future
28m8 Keys for Successful Serverless Architectures
18mServerless Best Practices
51mServerless – are you ready for the revolution?
01h 00mServerless Trends of the Season