Leadership
Decide where AI pays back, then make it happen.
At a glance
The board asks which AI work to fund, fix or stop.
- Deployed AI
Weeks of work
Managers collect and compile
- Operational AI
Hours each month
AI keeps the portfolio current; leaders decide
Illustration
What matters
What leadership teams are judged on.
Return on AI spend
Money moved from pilots to processes that pay back.
Risk and governance
Clear owners, limits and records the board can see.
Speed of decision
Board papers and reviews prepared in hours.
An organisation that keeps up
Teams trained to run AI in their own work.
Walkthrough
The same request, three ways.
The request
MessageChair of the board
Agenda item for the next meeting
Which AI work do we fund, fix or stop?
One paper, with owners and numbers.
Autonomy level 1
Deployed AI
A chat assistant. Someone asks, it answers, people do the rest.
- A person types
What are the best AI use cases for a mid-sized manufacturer?
- AI replies
Popular use cases include predictive maintenance, demand forecasting, quality inspection and customer service chatbots. Start with a pilot to prove value.
- Person
Strategy team. Spends weeks gathering numbers from each department for a slide deck.
3 steps · one to a person · weeks of work
- How long
- Weeks of work
- Who does the work
- Managers collect and compile
- What keeps it safe
- None: ideas without your numbers
- What goes to people
- Everything
Autonomy level 2
Specialised AI
It works from your data and tools and prepares the work. A person approves.
- Trigger
The quarterly AI review is on the board calendar.
- Tools
finance.costs("AI") · processes.measures(owners) · risk.register("AI")Returned: Nine AI initiatives. Three in daily use with measured savings; four pilots with no owner; two with open risk items.
- AI replies
A board paper with one line per initiative: owner, baseline, measure, monthly cost, risk and autonomy level, with fund, fix or stop proposed for each.
- Person
COO. Reviews the paper and changes two recommendations.
Edited - Action
Paper issued with the decisions requested.
5 steps · one to a person · two days of the COO's time
- How long
- Two days of the COO's time
- Who does the work
- AI compiles; leadership decides
- What keeps it safe
- Every figure linked to its source
- What goes to people
- Every decision
Autonomy level 3
Operational AI
It runs the process within set limits, logs every step and hands people the exceptions.
- Trigger
Monthly, from live process measures.
- Rule
Every AI process has an owner, a baseline, a measure and a budget.
Anything without an owner is flagged; anything over budget for two months is flagged.
- Action
Portfolio view updated; stalled pilots flagged to fix or stop.
- Person
Exception: goes to a person
Executive team. Two pilots have no owner, so the executive team decides: one is fixed with a named owner, one is stopped.
Approved - Record
[1st of the month]Decisions, owners and budgets kept in the AI register.
- Report
Quarterly to the board: return by process, open risks, and what was stopped.
6 steps · one to a person · hours each month
- How long
- Hours each month
- Who does the work
- AI keeps the portfolio current; leaders decide
- What keeps it safe
- A register, owners, budgets, board oversight
- What goes to people
- Every investment decision
Controls
- Owner
- Chief operating officer
- Measure
- Return per AI process
- Limit
- No process without an owner and a budget
- Log
- The AI register
- Fallback
- A manual quarterly review
- Stop
- Any owner can pause their process
Illustrative: names, figures and timings are invented.
Where to start
Processes that pay back first.
1Assist
An AI portfolio review
Where it pays back today, and where it never will.
1Assist
Process selection and business cases
Leaders choose; we bring the numbers.
2Automate with approval
AI governance and policy
Owners, limits and records.
2Automate with approval
Board and investor reporting
Drafted from live measures.
3Run with exceptions
A monthly AI operating rhythm
Reviews that prepare themselves.
2Automate with approval
Team training by department
Each team learns to run AI in its own work.
What we build and run
Built in your systems, run with your team.
- An AI portfolio review with a ranked plan
- Governance: owners, limits, a register and a policy
- A reporting rhythm built on live measures
- Team training by department
What we measure
Numbers you already track.
- Return per AI process
- Share of AI spend in daily use
- Pilots stopped or scaled
- Risks opened and closed
- Teams trained
The Academy
Learn to run it yourselves: the Leadership track.
Choose where AI pays back, govern it, and fund it like any other investment.
Being built