--- schema: 1 type: video title: 'Let''s integrate AI Agents in Event-Sourced Systems — Divakar Kumar, FlyersSoft' description: >- A card gets declined and no one, including the customer, can say exactly why. That gray zone is where Divakar Kumar points his agents. In a payments and fraud system, a rule based engine and an ML model already score most transactions cleanly; the hard cases are the ambiguous ones that neither can resolve. His approach adds an agentic layer on top of an existing event sourced architecture rather than replacing it, so the bounded contexts already in the system, transaction, device, and account, become the context the agents reason over. Events flow through change feeds into projections and a semantic layer that the agents read, communicating asynchronously through a message broker in a saga style loop. A risk analyzer agent, a second agent that reaches a verdict, and a third work the case while guarding against infinite loops and keeping memory short, all runnable serverless. The takeaway is architectural: event sourcing already carries the state and history an agent needs, so the cleanest way to add judgment to a production system is to layer agents onto the events you are already emitting. Speaker info: - https://www.linkedin.com/in/divakar-kumar/ - https://iamdivakarkumar.com Timestamps: 0:00 - Introduction: adding agents to an existing system 1:20 - A declined transaction you can't explain 3:04 - Where rule based and ML systems fall short 5:40 - Handling the gray zone with agents 5:53 - Bounded contexts: transaction, device, account 8:24 - Event sourcing and change feeds 10:57 - Building the semantic layer 13:16 - Avoiding infinite loops 14:07 - The risk analyzer and verdict agents 15:24 - The saga orchestration loop 19:00 - Putting the architecture together canonical: >- https://liquidgold.cc/channel/aidotengineer/explore/video/let-s-integrate-ai-agents-in-event-sourced-systems-a8d8e7740622 markdown: >- https://liquidgold.cc/channel/aidotengineer/video/let-s-integrate-ai-agents-in-event-sourced-systems-a8d8e7740622.md llms: null parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T06:29:41.826Z' language: null children: [] videoId: o6U_2vd967Y durationSeconds: 1297 publishedAt: '2026-07-30T00:00:06.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 23 endSeconds: 90 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=23' heldReason: rights - ordinal: 1 text: null startSeconds: 198 endSeconds: 349 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=198' heldReason: rights - ordinal: 2 text: null startSeconds: 100 endSeconds: 188 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=100' heldReason: rights - ordinal: 3 text: null startSeconds: 198 endSeconds: 349 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=198' heldReason: rights - ordinal: 4 text: null startSeconds: 371 endSeconds: 512 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=371' heldReason: rights - ordinal: 5 text: null startSeconds: 371 endSeconds: 512 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=371' heldReason: rights - ordinal: 6 text: null startSeconds: 523 endSeconds: 658 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=523' heldReason: rights - ordinal: 7 text: null startSeconds: 523 endSeconds: 658 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=523' heldReason: rights - ordinal: 8 text: null startSeconds: 658 endSeconds: 670 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=658' heldReason: rights - ordinal: 9 text: null startSeconds: 670 endSeconds: 802 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=670' heldReason: rights - ordinal: 10 text: null startSeconds: 802 endSeconds: 814 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=802' heldReason: rights - ordinal: 11 text: null startSeconds: 861 endSeconds: 929 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=861' heldReason: rights - ordinal: 12 text: null startSeconds: 861 endSeconds: 929 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=861' heldReason: rights - ordinal: 13 text: null startSeconds: 941 endSeconds: 1144 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=941' heldReason: rights - ordinal: 14 text: null startSeconds: 670 endSeconds: 802 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=670' heldReason: rights - ordinal: 15 text: null startSeconds: 941 endSeconds: 1144 watch: 'https://liquidgold.cc/watch/o6U_2vd967Y?t=941' heldReason: rights --- # Let's integrate AI Agents in Event-Sourced Systems — Divakar Kumar, FlyersSoft A card gets declined and no one, including the customer, can say exactly why. That gray zone is where Divakar Kumar points his agents. In a payments and fraud system, a rule based engine and an ML model already score most transactions cleanly; the hard cases are the ambiguous ones that neither can resolve. His approach adds an agentic layer on top of an existing event sourced architecture rather than replacing it, so the bounded contexts already in the system, transaction, device, and account, become the context the agents reason over. Events flow through change feeds into projections and a semantic layer that the agents read, communicating asynchronously through a message broker in a saga style loop. A risk analyzer agent, a second agent that reaches a verdict, and a third work the case while guarding against infinite loops and keeping memory short, all runnable serverless. The takeaway is architectural: event sourcing already carries the state and history an agent needs, so the cleanest way to add judgment to a production system is to layer agents onto the events you are already emitting. Speaker info: - https://www.linkedin.com/in/divakar-kumar/ - https://iamdivakarkumar.com Timestamps: 0:00 - Introduction: adding agents to an existing system 1:20 - A declined transaction you can't explain 3:04 - Where rule based and ML systems fall short 5:40 - Handling the gray zone with agents 5:53 - Bounded contexts: transaction, device, account 8:24 - Event sourcing and change feeds 10:57 - Building the semantic layer 13:16 - Avoiding infinite loops 14:07 - The risk analyzer and verdict agents 15:24 - The saga orchestration loop 19:00 - Putting the architecture together Canonical: https://liquidgold.cc/channel/aidotengineer/explore/video/let-s-integrate-ai-agents-in-event-sourced-systems-a8d8e7740622 · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.