Start here

This site is a working record of one question: what it actually takes to turn a new technology into something a large, regulated organisation can govern, fund, staff and rely on.

The question is worth the trouble because it is usually where the value lands. In every technology shift, a large share of the return has gone not to whoever invented the thing but to whoever worked out how institutions could use it safely and at scale. That work is unglamorous, badly documented, and mostly learned by getting it wrong first. This is my attempt to write some of it down.

All of it comes from doing the work rather than advising on it, and all of it is generalised on purpose. No confidential detail from any employer, no invented numbers, no case study with a logo at the bottom.

What is on the site

Frameworks is the load bearing part. Four working models for problems I keep meeting: why proofs of concept stall, where a platform’s responsibility should end, how to make governance a route rather than a wall, and how to prioritise when every function wants something different and all of them are right. Each one states where it breaks.

Selected work is three generalised accounts of real problems, written to show the reasoning rather than the result.

Writing is shorter and more argumentative. It is where an idea gets tested before it is settled enough to be a framework.

Books is the shelf underneath all of it, with the one idea I take from each and the point where that idea stops being useful.

Where to go first

• If you have a proof of concept that works and cannot move: The Production Distance Model.
• If you are arguing about what your platform should own: The Platform Responsibility Line.
• If a control function is the thing in your way: Governance as a Path.
• If you are deciding whether any of this is worth your time: open any essay in Writing. If the first two paragraphs do not tell you something you did not already know, the rest will not either.

What this site will not do

• It will not disclose anything confidential about any employer, past or present, and nothing here represents an employer’s views.
• It will not sell you a maturity model, a certification or a transformation.
• It will not claim these frameworks are validated. They are working models, drawn from one kind of organisation, and they say so.

Who is writing this

Chad Karannagoda. I work in enterprise technology in New Zealand, currently as the service owner of a shared machine learning platform inside a regulated bank. Engineering first, then product, which is why most of what I write is about the distance between the two. The longer version is on the About page, and the fastest way to reach me is Contact.