Recent content on Filippo Diotalevi
For a decade my advice to Engineering leaders was to stay out of the codebase and scale through people. AI changed that: a leader can now affect productivity and quality directly, with their own hands. Not by picking up production tickets, but by prototyping ahead of the team, strengthening the tooling everyone relies on, and building agents that continuously improve the organisation.
When we move from AI-aided software production to Governance of AI-enhanced software production, things get less clear. Mistakes get expensive and visible, but there are still no established frameworks to manage it. I look at the three areas I find most critical right now: security, cost, and audit defensibility.
My name is Filippo Diotalevi . I write about software engineering leadership, trying to make sense of more than two decades working in corporates and startups through hypergrowth and hiring freezes, greenfield rewrites and legacy migrations, reorgs, lots of failures and the occasional and rare big success. 
 Elsewhere 
 Probably the most reliable way to reach me is sending a message via…
As AI reduces the cost of writing code, different skills become fundamental for Engineering Managers and CTOs. None of them is new. AI just makes failing at them visible — and existential.