AI-Native Software Economics Board Adviser

The mandate

The board keeps returning to a question its subscription history cannot answer: what should the company charge when an engineering agent completes a work package rather than helping a licensed user perform one? Seat expansion, generated artefacts, compute consumption and claimed hours saved tell conflicting stories, while retries, human correction and contractual responsibility can turn apparent automation success into an unmeasured service obligation.

The adviser will commit three working days each month: one for cohort and cost evidence review, one for management challenge, and one split between the chair and product-finance scenario work. Five Toronto Finance and Product Committee meetings and quarterly Waterloo workshops are included; a written view on a time-sensitive pricing or investment paper is due within two Canadian business days, with unscheduled support capped at eight hours monthly.

The first term runs for nine months from 5 January 2027. During month seven, the chair will lead an effectiveness and conflict review, after which the full board may approve one extension of no more than three months; neither management nor an investor director may renew the appointment alone or change the equity terms retrospectively.

This is an influence-only board appointment with no line authority or executive responsibility. The adviser cannot set a price, approve discounting, direct product releases, accept customer liability, hire staff or represent the company to investors; accountable executives make those choices and the committee minutes where advice, evidence and final judgement diverge.

No more than two other material board or advisory commitments may be held during the term. Work for a competing engineering-agent platform, a pricing vendor under consideration, a customer negotiating outcome-linked terms or an investor with a live position in the company requires prior disclosure and may demand recusal; use of confidential cohort economics elsewhere ends the appointment.

Why the board wants this voice

The directors understand conventional recurring revenue but have not scaled a product whose autonomous work can reduce the very user count on which historic pricing depends. Product leaders favour adoption signals, Finance focuses on gross margin and Sales reports booked expansion, leaving no shared test for economic quality. The chair wants an operator who has seen AI capability alter value metrics, customer liability and software multiples at the same time.

What you will own

• Challenge whether seats, agent runs, accepted engineering outputs, cycle-time reduction or risk transferred to the vendor is the defensible unit of customer value.
• Test cohort evidence for expansion by separating durable workflow penetration from promotional credits, parallel manual effort, implementation labour and correction burden.
• Press management to expose retry compute, review effort, exception handling, support escalation and contractual remediation inside every proposed pricing architecture.
• Shape guardrails for outcome-linked commitments so acceptance definitions, customer dependencies, quality thresholds and liability caps remain commercially legible.
• Probe how an agent that reduces licensed users should change retention measures, sales incentives, revenue forecasting and the board’s interpretation of net expansion.
• Compare packaging choices across individual copilots, pooled autonomous capacity and completed-work bundles without turning external anecdotes into unsupported benchmarks.
• Frame the committee’s decision questions for financing materials, ensuring growth claims reconcile to cash collection, contribution quality and observable customer value.

Candidate qualifications

• Served a software board or executive committee while an AI-enabled product moved from user-assistance pricing toward autonomous or outcome-linked commercial models.
• Redesigned packaging after automation reduced seat demand and can evidence the resulting changes in adoption, retention, margin and sales behaviour.
• Diagnosed customer-level economics including inference consumption, implementation labour, human review, retries, support burden and contractual credits.
• Challenged a growth metric that overstated durable expansion and helped directors adopt a measure that later reconciled to renewal and cash evidence.
• Negotiated or governed commercial terms where software accepted responsibility for completed work, requiring clear quality, dependency and liability boundaries.
• Managed portfolio-board conflicts involving investors, direct competitors, pricing advisers and enterprise customers without leaking comparative operating data.

Non-negotiables

• Can protect three days every month, attend all five Toronto committee meetings in person and travel quarterly to Waterloo for the complete nine-month term.
• Will disclose current board seats, investments, retainers and customer relationships before receiving company, cohort or financing information.
• Accepts that the option grant is illiquid, subject to dilution and approval, and cannot substitute speculative equity value for the stated cash retainer.
• Has directly changed AI-software monetisation and measured post-launch customer economics; general SaaS governance experience alone will not qualify.

1. 49 words maximum. Which current company, investor, customer or vendor relationships would need disclosure to this board, and what recusal would you propose?
2. 49 words maximum. Describe one AI product whose value metric you changed after automation disrupted seat economics, including the evidence observed after launch.
3. 49 words maximum. Confirm the Toronto and Waterloo cadence and explain how you would test whether outcome pricing conceals human correction or customer-liability cost.