Who is Stefano?

Stefano Milanesi is an Agile practitioner, Professional Scrum Trainer (PST), and Professional Kanban Trainer (PKT) with more than 25 years of experience in the software industry.

He works with product teams to help them achieve better outcomes through practical, human-driven ways of working. His focus is not only on processes and frameworks, but also on creating workplaces where people can take ownership, collaborate effectively, and solve complex problems.

Stefano began his career in medical physics and software engineering, working with neural networks in diagnostic imaging. He later founded and led several successful software companies. His experience as a CEO and entrepreneur strengthened his belief that effective organizations are built around people, not rigid processes.

Today, he works as a trainer, coach, and consultant, helping teams and organizations apply Scrum and Kanban in a practical and meaningful way. He is also a regular speaker at international conferences, where he shares his experience, challenges common assumptions, and learns from the wider agile community.

Faster Developers, same long wait - Making Flow visible to show where AI actually helps

My client was a fintech with a Time-to-Market problem. The managers did the sensible thing. They planned to hire twenty people to make the six Scrum teams faster.

The teams got slower.

So they tried AI. They funded training for the developers, and this time something worked. Code got written faster. Defects went down. The developers were genuinely happier. Everyone expected the product to start reaching customers sooner.

It didn’t. Time-to-Market stayed flat. For some features, the wait got longer.

Here is the uncomfortable truth we had to put in front of leadership. Writing code is a small slice of how a product reaches a customer. We had made that slice faster — almost twice — while work kept piling up. It waited for code reviews. It waited on other teams. It waited for someone with the authority to say yes. Without flow, every effort to improve the product is wasted motion.

This is the story of what we did next. We used Kanban to make those queues visible to the people funding the next decision. I will show you the conversations that followed — including the moment leadership stopped reading developer speed as system progress.

And I will share the twist. The thing that finally moved Time-to-Market was pointing AI not only at the developers, but at the bottlenecks outside development — freeing Product Owners to decide faster.

I'm not going to hand you a solution. I'm going to show you how we found ours, starting from the one question that started it: where does the work actually wait? You'll leave with elements that can help you think about where to focus, and that could include your own AI investment.

Three key takeaways:

1. Spot where work actually waits in your value stream — review, cross-team dependencies, decision and approval queues — and articulate to leadership why faster development alone won’t improve flow or Time-to-Market.

2. Apply a Kanban diagnostic — Definition of Workflow, Work Item Age, and WIP limits — to make a hidden bottleneck transparent to the people who fund decisions.

3. Leave with one concrete experiment to surface a queue in your own process and decide where to point your next investment — including where to aim AI.