Sticky API: Connect Bricks and enter a global IT-ecosystem
summary
Argues that a production API must be built as modular services with automated tooling: Spryker uses JSON:API but allows multiple formats, a separated API application plus format libraries, and code generators to produce IDE-ready endpoints. Implements ACL, scopes and field-level permissions, pushes normalized data asynchronously into Redis and Elasticsearch, forbids ad-hoc cache keys in favor of incremental invalidation, and generates Swagger docs directly from code so documentation never drifts. Claims this enables parallel team releases and handles sudden load jumps (example: 10 req/min to 81k reqs) without breaking integrations.
AI-written summary · about
chapters (24)
- 0:00Intro
- 1:12Technology Evolution
- 2:10Single Source
- 2:54Targeting Omni-channel
- 4:29Communication
- 6:26External Ecosystem
- 7:34Global Ecosystem
- 8:34Daily Challenges
- 11:46Data Control
- 15:21New Development process
- 16:16Code Structure
- 17:12Code generators
- 17:59API Maturity
- 19:21API Standards
- 20:16API Formats
- 20:39Storefront vs Backoffice
- 22:24Spryker Glue API
- 24:58Scalability
- 25:22Data Cache
- 26:32Spryker Architecture
- 27:21Vendor Versioning
- 28:51Reusable Resources
- 31:55Glue API Documentation Generator
- 32:53Take aways
related talks
23mWhat Makes an Agile API?
14mHow to build better and consistent APIs at scale
27mGetting developers hooked on your API
25mAPIs for Scaling Enterprises: Setting and Applying API Standards.
37mAutomation++ - Building an Enterprise API Economy from scratch
16mAPI Integration with JS: Scalable, and Future-Ready Systems
28mThe New API Stack
21mPragmatic Strategies for Global API Integration