Abstract

Improving product and flow in organisations sounds simple. In reality, it means answering the same questions repeatedly, correcting misconceptions, translating theory into practice, and being “always on” for every team. People say new ways of working are too theoretical. Or they don’t know where to start. Or they can’t find the right information. Or they interpret the same metric five different ways.

In this talk, I’ll show how I tackled that problem by building an AI agent: the Product & Flow Advisor.

This agent combines delivery theory (flow metrics, probabilistic forecasting, outcome thinking, prioritisation models) with live organisational data and internal documentation. It answers questions consistently, surfaces risks, accelerates analysis and reduces the dependency on a single expert being available at all times. This is not a theoretical AI talk. It’s a practical case study of augmenting product and flow leadership with an agent that scales knowledge, reduces friction, and shortens feedback loops.

I’ll demonstrate how the agent helps teams to:

• Get clear, context-aware answers about ways of working
• Apply prioritisation scoring models correctly and quickly
• Define outcome-based initiatives with stronger thinking upfront
• Understand team performance using flow metrics
• Spot delivery risk, aging work, and red RAG projects early
• Move from opinion-driven debate to evidence-guided discussion

You’ll leave with clear ideas on:

• Where AI agents can genuinely add value in product/delivery environments
• How to structure knowledge so agents give high-quality answers
• The risks of naïve AI adoption in delivery contexts
• How to start small and build an internal advisor that people actually use