Platforms: Build abstractions, not illusions
summary
Platforms claim to reduce cognitive load but real success hinges on clear abstractions, not hidden defaults or illusions like RPC that pretend network calls are local. Concrete signs of a working platform: someone builds something that surprises the team (a CICD pipeline repurposed as a data delivery system), a product exposes business-aligned services such as a database with customer data that cloud providers won’t offer, and pricing still makes sense (cloud servers at ~€0.50/hr). Think fruit salad vs fruit basket and name features by purpose — accelerator not gas pedal.
AI-written summary · about
chapters (13)
- 0:00The challenge of complexity
- 2:41Defining platform value
- 13:49Testing your platform
- 16:04The platform meta-truth
- 18:45Removing IT constraints
- 22:30Standards as enablers
- 27:33Dynamics of platforms
- 31:44Organizational maneuvers
- 34:40Abstraction versus illusion
- 39:08Designing for domains
- 46:35Metaphors for strategy
- 51:22Human expertise and AI
- 56:09Final platform advice
related talks
46mPlatforms: Build abstractions, not illusions
46mPlatforms: Build Abstractions, not Illusions
34mPlatforms: Defying the laws of IT physics
51mThe Magic of Platforms
48mBuild Abstractions Not Illusions
29mPlatforms, Architecture and Sustainability
23mPlatform Engineering for Architects - Crafting Platforms as a Product
32mPlatform Engineering for Software Developers and Architects (Redux)