AI & ENGINEERING / THE WORK
What happens when AI becomes part of engineering?
I’m less interested in predicting the future of software engineering than in experimenting with it. This is where I explore what changes when AI becomes part of the way engineering organizations actually work.
A DIFFERENT QUESTION
The obvious question is: How much faster can AI make engineers?
The more interesting question, for me, is what happens to the system around the engineer when implementation becomes dramatically easier.
Requirements. Product discovery. Development. Testing. Delivery. Support. Continuous improvement. AI can potentially reduce friction between all of these — not just accelerate one step in isolation.
WHAT I’M EXPLORING
AI across the software lifecycle.
Requirements
Can AI help Product Owners turn ideas into clearer, more complete and more realistic requirements by understanding the organization’s own documentation, backlog and code?
Development
What changes when engineers work alongside coding agents and AI becomes part of everyday implementation rather than an occasional assistant?
Delivery
If implementation gets faster, where do the real constraints move? I’m interested in architecture, testing, dependencies, environments and organizational coordination.
Support & improvement
Can an organization turn its own support history into knowledge — and then use that knowledge to reduce recurring problems?
AI AS AN ORGANIZATIONAL BRIDGE
The most interesting AI tools may sit between teams.
One of my current experiments connects AI to JIRA, Confluence, GitHub and the filesystem through MCP. The aim is not simply to generate text. It is to give AI enough organizational context to help translate ideas into realistic delivery work.
That creates a different kind of engineering opportunity: using AI to reduce the gaps between Product, Engineering and the systems they depend on.
ENGINEERING LEADERSHIP
AI changes the tools quickly. Organizations change more slowly.
That gap is where engineering leadership becomes interesting. How should teams work? How do responsibilities change? What happens to quality, architecture, career development and decision-making? And where does human judgement become more important as AI becomes more capable?
THE METHOD
Discover → Experiment → Reflect → Share.
I build things, try tools, look at what actually happens, and document the useful parts — including what doesn’t work. The goal is practical learning, not AI theatre.