Operating Systems as a Service: The Evolution of Platforms
summary
Argues that enterprises need domain, life and work operating systems built from a core events-and-functions layer, not ad hoc stacks; defines five concrete subsystems—sensors, system of records, system of intelligence, system of interaction, system of operations—and shows workflows (execution, exception, expansion, efficiency) that stitch them together. Gives concrete builds: speech-to-text + text-to-intent + knowledge graph + feature store + drones for an agricultural advisory workflow, human-curated AI loops for clinical diagnosis, Kafka-style event queues tying hundreds of models into realtime decisions.
AI-written summary · about
chapters (9)
- 0:00Operating systems as a service
- 3:38Three OS needs
- 10:18Ecosystem product complexity
- 14:08Breaking systems into workflows
- 22:34AI-embedded workflows
- 29:12Augmented intelligence in practice
- 35:20Emergent intelligence frontier
- 39:32OS as a service architecture
- 42:21Path to intelligent organizations
related talks
34mPlatforms: Defying the laws of IT physics
14mModern Services Demand a DevOps Culture Beyond Apps
51mThe Magic of Platforms
48mPlatform Engineering is Domain Driven Design
24mBOSS: Building Operating System Services
47mPlatform Engineering is Domain-Driven Design
40mSystems Thinking for Developers
58mPlatforms: Build abstractions, not illusions