Autonomy is table stakes. The moat is everything around the agent.
Real builds on the agentic stack. Each episode takes one layer everyone is talking about (agent loops, generative UI, agents that pay), turns it into a working system, then ships the film and the code. These aren't toys: every build carries the parts that decide whether agents survive production, namely verification, orchestration, human oversight, grounding, and an audit trail.
- Essays, one per layer: zishanalikhan.com/series/agents-orchestrated
- Visual companion, with the films, the repos, and an interactive codemap of all three codebases: zishanalikhan.com/agents-orchestrated
That index is generated from the source, so it is always current. This README deliberately does not repeat it.
| Episode | Repo |
|---|---|
| EP 01 · Loops | what-are-loops |
| EP 02 · AG-UI | camunda-agui-wealth-demo |
| EP 03 · AP2 | camunda-ap2-payments-demo |
Every build is public and runnable.
Zishan Ali Khan, Enterprise AI GTM & Solutions at Camunda. My day job is putting agents inside real business processes; this series is where I build each layer end to end and publish what holds up.