Workflow ecology: what sits underneath every step of a process
Progentic — https://progentic.co.nz/videos/workflow-ecology
Published 17 August 2026 · 1:01

Transcript

[0:00] Only about ten percent of organisations fully understand the workflows which create value for their customers. So the first major part of any automation or AI project is fully describing the workflows.

[0:13] If you place an AI or automation solution without understanding the surrounding dependencies, for example, you risk solving the wrong problem or designing the wrong solution.

[0:24] Understanding a process flow is not a two-dimensional job. Every step has an entire ecology below it.

[0:31] Data, data flows, people, roles, permissions, communications, measurements. A workflow is like a building, with a hive of activity under every step. And that ecology changes as you move through the process. Layers in one step reach into layers in another.

[0:50] The job of a workflow engineer is to lift the lid, and we build models like this in many of our engagements, because that's where the AI and automation opportunities actually live.

Key points

- Only about one in ten organisations can fully describe the workflows that create value for their customers, so describing them is the first real task in any automation or AI project.
- A process map is not two-dimensional. Under every step sits data and data flows, people, roles and permissions, communications, documents, systems, policy and regulation, customers and KPIs.
- That ecology changes as you move through the process, and layers in one step reach into layers in another. The dependencies are exactly what a flat diagram hides.
- Place an automation or AI solution without understanding those dependencies and you risk solving the wrong problem, or solving the right one the wrong way.
- The opportunities live in the layers rather than in the boxes, which is why we build layered workflow models in most engagements.