Diligence · Operating decision guide

What should a portfolio-company board ask about AI?

Otomat editorial · Revised 5 November 2026

Ask for the operating decision, capital committed, evidence achieved and the next choice. A board needs a clear continue, change or stop decision more than a count of pilots.

Put the decision at the top

Each initiative should identify a business constraint, an accountable executive and a decision the board or delegated committee can make. Separate the amount already spent from the next requested commitment. A slide showing many use cases hides whether management is seeking approval, reporting an accepted result or asking for help with a blocked dependency.

Keep evidence stages visible

Use distinct labels for an idea, an evaluated candidate, a bounded live workflow and a finance-reviewed outcome. Describe the acceptance evidence behind the label. A technically working prototype may still lack staff capacity, use rights or a tested fallback. Conversely, a modest workflow with an owner and a repeatable benefit may deserve support without an ambitious platform story.

A board decision cover sheet
Check	Evidence to bring	Decision or stop condition
Capital request	Spent to date, next commitment and reversible scope	Approve the next gate, not an unspecified transformation budget.
Evidence stage	Result, observation window and acceptance owner	Do not describe a prototype as a realized financial outcome.
Management capacity	Combined hours needed from scarce owners and reviewers	Sequence work when concurrent commitments exceed capacity.
Continue or stop	Threshold, unresolved blocker and decision date	Require a named decision instead of another status-only meeting.

Ask about the management constraint

Check which executive, operator and reviewer must contribute time. If five projects depend on the same controller, the board should see the combined demand, not five individually small requests. Include unresolved security, data or supplier issues and the person who can resolve them. The sponsor’s enthusiasm does not create operating capacity inside the portfolio company.

Make stopping an acceptable answer

Define what new evidence would justify continuing, changing scope or ending the work. Preserve what was learned and the reusable artifacts if a project stops. Reporting an unsuccessful test honestly can protect capital and management attention. Do not combine projected benefits across initiatives when they depend on the same hours, customers or cost line.

Questions before the decision

Should every pilot go to the board?

Use the company’s delegated authority. Escalate commitments, risks and unresolved decisions that exceed that authority; keep a consistent evidence record for the rest.

What belongs in the headline?

The decision needed, business owner, evidence stage, next commitment and principal uncertainty. Put model selection details in supporting material unless they change the decision.

Take the evidence sheet to the meeting

Download the blank CSV, assign an owner to each check and record the evidence. Keep completed sheets in your own approved system.

Download the evidence sheet (CSV)

Use the working tools

Blank templates are ungated. Adapt them to the workflow; inputs are not submitted.

Open the operating charter

Sources and limits

These are Otomat’s proposed operating methods and original decision artifacts. External references inform risk management; they do not certify this method or establish client results. Linked sources were selected for this review; their live revision dates may change.

For anonymous documentation and proposal patterns, see delivery evidence and its boundaries.

Related decisions

Bring a workflow to the discussion

Share the business goal and constraints through the mandate process. Do not send sensitive production data in the initial brief.