Micro Minds: Event-Driven Services Modeled as Neural Synaptic Networks
summary
Argues that event-driven microservices can be designed like neural networks, wired as simple entities/actions/views that emit and transform events into commands; demonstrates the Earthship demo generating about 10,000 entities and ~15,000 event lines from one HTTP request, proving the choreography Saga, one‑to‑many and reduction tree patterns scale and work under heavy load. Uses Java SDKs, Blender for the animation and AI assistants like GitHub Copilot and GPT‑4 to write code; relies on at‑least‑once delivery and idempotent handlers.
AI-written summary · about
chapters (15)
- 0:00Micro Minds introduction
- 1:12AI and event-driven inspiration
- 3:00Micro Minds and biology
- 4:04Pure event-driven design
- 9:25Live event-driven demo
- 11:11Event-driven patterns
- 14:20Code walkthrough
- 16:51AI assistants in dev
- 20:01Earthship demo
- 24:38Pure event-driven data
- 27:28Entity patterns deep dive
- 31:48Idempotency in events
- 35:41Testing event-driven services
- 37:29Future of prompt-driven code
- 40:18Key takeaways
related talks
54mMicro-Mind Services: Building Complex Behavioral Systems
36mEvent-Driven Services as Neural Synaptic Networks
48mYou need Event Mesh, not Service Mesh
40mCommunication between (micro-)services
47mEvent-Driven Microservices = Microservices Done Right
54mHow Events Are Reshaping Modern Systems
58mImplementing Event-driven Microservices
29mDesigning Microservice Architectures