AI Decision Review Notes

AI Decision Review Notes

Practical writing for leaders making AI decisions

Practical writing for leaders making AI decisions

Frameworks, questions, and operating models for turning high-stakes AI initiatives into reviewable decisions leaders can trust.

Frameworks, questions, and operating models for turning high-stakes AI initiatives into reviewable decisions leaders can trust.

AI governance

Board strategy

AI risk

LATEST WRITING

Start with the decision layer

Professional reviewing documents beside a laptop

Exit readiness & due diligence

AI made the standard part of due diligence cheap. Why haven’t fees moved?

DD vendors now run the standard checklist with AI, yet fees stay the same. What that means for how PE funds brief and pay for assessments.

3 min read

Team meeting around a wooden table

AI strategy & operating model

PE teams have brilliant AI users and a fund that barely benefits

Associates build their own Claude skills, yet each one stays with the person who built it. Two steps to get AI value at fund and portfolio level.

3 min read

Leadership team discussing plans at a table

Business cases & ROI

Is AI an efficiency play or a growth play?

PwC's CEO survey shows most companies see AI on the cost side. Most initiatives have room for both, if you define how saved time becomes money before you build.

3 min read

Office desk with notes and digital work context

AI strategy & operating model

Two years into AI adoption, still starting from zero context

A fast-moving company with its own chatbot and AI ambassadors still has most AI use starting from scratch. Why missing organisational context caps your AI value.

3 min read

Analytics dashboard on a laptop screen

AI costs & infrastructure

AI gets cheaper every quarter. Why do the bills keep rising?

Unit costs are falling fast, yet AI bills go up. Three reasons why, and a three-layer way to measure AI use without tagging every conversation.

3 min read

Person working through documents at a cluttered desk

Workflow & data

Two rejected applications, two completely different fixes

One application missed a signature. Another failed a rule a near-identical case had passed. Why these need completely different fixes.

3 min read

Computer screen with code and AI action menu

AI strategy & operating model

What 'agentic' used to mean, and why workflows pay off first

Most of what gets sold as agentic is a single LLM call. Even real agents are rarely where the ROI is today. Where workflows you can test do the job better.

3 min read

Network switch and connected server cables

AI costs & infrastructure

What sovereign AI actually means

Choosing the EU region of a US cloud doesn't make you sovereign. Three questions that define sovereign AI and three places you can land.

3 min read

Hands typing on a laptop with spreadsheet data

Business cases & ROI

From gross savings to net: where GenAI business cases erode

Gross savings rarely become net ROI. The four cost lines that erode a GenAI business case, with practitioner ranges for each.

3 min read

Laptop on a clean desk for technology strategy work

Exit readiness & due diligence

Where your AI advantage actually comes from

Funded teams can rebuild AI features in weeks. Where a real AI moat forms, and three questions to find out where you stand.

3 min read

Documentation and process notes on a screen

Workflow & data

Zero-based redesign: what a GenAI compliance project taught me

In a compliance project, we stopped asking how to review faster and asked whether humans should review first at all. The results changed the work.

3 min read

Strategy matrix with sticky notes on a table

Business cases & ROI

Kill, Rescope, Alter or Scale: triaging your GenAI portfolio

Most GenAI portfolios carry initiatives that should have been stopped months ago. A matrix to decide what to kill, rescope, alter or scale.

3 min read

Dornovo Copyright 2025

christan@dornovo.com

Dornovo Copyright 2025

christan@dornovo.com

Dornovo Copyright 2025

christan@dornovo.com