Kevin Byrne-Langlois

Can a regulated company trust software built by AI?

I'm a CTO in regulated software. Management keeps asking me what AI can really do, and whether we can stay in control of it. So I'm finding out, in public.

AI agents build a real product across the whole lifecycle, under the controls a regulated company would demand. I act as the CTO: requirements, risk, review and decisions. Every two weeks I publish what the numbers say, failures included.

The questions

The ones I hear from management in regulated industries. More will come.

  1. A

    Can AI produce enterprise-level code, and is there technical debt in that code?

    Open
  2. B

    Can anybody now build an enterprise-level application, and so become a competitor, or a danger?

    Open
  3. C

    Is it possible, today, to fully automate the software development lifecycle with AI?

    Open
  4. D

    Can you keep control and visibility over an AI-automated pipeline?

    Open

The product under test tackles a problem of its own: control and visibility over the actions, costs and impacts of all AI used in a regulated enterprise.

Latest posts

The first post goes up at the end of the first cycle.