Event-Driven Evolution with AI as my Partner in Design
summary
Legacy Laravel application evolves to an event‑sourced model using Gemini as a design partner, producing concrete decisions: emit events to AWS EventBridge then Kinesis Firehose to S3 as the event store, document payloads with JSON Schema, and enforce a transactional outbox so DB updates and event writes stay consistent. Financials model uses integer money movement legs (multiple debits/credits per event), allocation/settlement event types, and example projections validated against real £1bn+ sales.
AI-written summary · about
chapters (31)
- 0:00Legacy monolith challenges
- 2:30Tillow and AI in design
- 4:58Event sourcing with AI
- 5:20Tillow gift card model
- 7:52Financial model problems
- 8:34Event sourcing benefits
- 9:16Event storming
- 10:40Domain discovery with AI
- 11:44AI transcription benefits
- 12:24Learning financial concepts
- 13:08Financial journals explained
- 15:55Business and technical domains
- 16:57Event streaming technologies
- 17:39Choosing AWS EventBridge
- 18:41Laravel to EventBridge
- 19:03Event structure design
- 20:06Tillow event envelope
- 22:15Data and metadata separation
- 23:22Journal entries as events
- 24:03Double entry accounting
- 25:50Complex scenarios with AI
- 26:56AI validation of projections
- 27:38AI naming conventions
- 28:41Event JSON schema
- 29:03Event types code
- 32:36Allocation and settlement patterns
- 34:02Outbox pattern for consistency
- 35:29Laravel event builder
- 37:38AI's role in development
- 39:25Human craft in software
- 40:52Q&A
related talks
46mShifting Gears: From Events to Event-Driven
43mEvent driven development
54mThinking in Events
01h 01mLegacy Architecture Migration Patterns with DDD
56mDesigning Event-Driven Applications to Improve Scalability
41mAccelerating Domain-Driven Design with AI
43mUnraveling the Threads: Event-Sourcing, CQRS, and Event-Driven Architecture
59mOptimizing Complex Workflows with Event - Driven - Multi Agentic Approach