Streaming Data Events Into An AI Cognitive Event Mesh Using AsyncAPI
summary
Shows live streaming from Oracle GoldenGate into a Solace event mesh via AsyncAPI, capturing DB commits into trails and creating AsyncAPI channels that preserve source order and provide exactly-once semantics. Demo builds a RAG-style AI connector that ingests orders and shipping events into a context DB and answers natural‑language queries: it reports 82 unique orders, with 27 today versus 11 yesterday. Implementation uses WebSockets, JSON Schema metadata for evolving schemas, and working request‑reply chatbot integration.
AI-written summary · about
chapters (18)
- 0:00Streaming data events with AsyncAPI
- 1:55Point-to-point sync integration problems
- 4:24Event-driven architecture for real-time engagement
- 8:15Decentralized event mesh and taxonomy
- 10:43AI data requirements
- 12:08Demo: AI-powered context mesh
- 14:13Event mesh with Solace and AsyncAPI
- 19:55Oracle GoldenGate intro
- 20:38Challenges in data-driven apps
- 23:08What is Oracle GoldenGate
- 25:37GoldenGate for hybrid/multicloud
- 27:24GoldenGate Data Streams and AsyncAPI
- 28:48Source order consistency with GoldenGate
- 31:35Exactly-once semantics in AsyncAPI
- 36:31Schema management for consumers
- 39:41Live: GoldenGate with Solace
- 44:22Demo: AI chatbot with real-time data
- 50:08Conclusion and Q&A
related talks
29mAgentic AI Meets Event-Driven Architecture
27mDistributed AI Processing with GoldenGate 23ai
35mMaking Open Banking Responsive & Agile with Asynchronous and Event-Driven APIs
18m3 Oracle AI Use Cases
51mEvent Driven AI: Supercharging ChatGPT with RAG & LangStream
46mState or events? Which shall I keep?
42mAgentic AI with AWS Streaming
32mEvent-Driven & Multi-Agent Architectures for Gen AI Workflows