An 8-to-16 week multi-stakeholder engagement that takes a firm from AI strategy to a funded build plan with a named owner.
For finance firms past the assessment phase that need an operating model around AI, not another deck. $25,000–$50,000. Fixed scope, fixed fee, signed-off deliverables at each phase.
The problem this service solves
A firm that has been through an assessment, or has done internal discovery on its own, faces a different problem: the assessment surfaced where AI pays back, but nobody has the time to design the operating model around it. Who owns the build? How does the second line review the work? What is the governance frame? What is the data architecture? What is the change-management plan for the operations team whose jobs are about to shift? Strategy without an operating model is a binder on a shelf. The engagement designs the operating model — staffing, governance, data flow, vendor stack, and rollout sequence — so the firm can actually execute.
What we deliver
Design
The AI operating model for the firm — roles, reporting lines, second-line review integration, and the governance committee charter. Sized to the firm's existing structure, not a McKinsey-deck org chart.
Specify
The data architecture changes required to support the top 3-5 priority workflows. Named systems, named integrations, named owners.
Build
The 12-month rollout plan with phase-gates, fundable line items, and a named accountable executive at each phase. Each phase has a kill condition.
Draft
The AI governance policy and the model risk management addendum, scoped to your existing risk framework (SR 11-7 if banking, NAIC #880 if insurance, fiduciary if RIA).
Brief
The board's risk and audit committees on the AI operating model and governance, with the materials they will need to oversee the program quarterly.
Sequence
Vendor and build/buy decisions for the priority workflows — including the explicit not-now list with the conditions that would change the answer.
How we work
Strategy alignment
Weeks 1-3Stakeholder interviews across executive, line-of-business, risk, compliance, and IT. Output: a strategy document the executive team signs off on before any operating-model design starts.
Operating model design
Weeks 4-10Roles, governance, data, vendor stack, and rollout phasing. Working sessions every two weeks with the project sponsor and core stakeholder group. Output: a signed-off operating model document.
Build plan and handoff
Weeks 11-1612-month rollout plan with phase-gates, owner per phase, kill conditions, and quarterly board-reporting cadence. Output: the build plan, the governance package, and a 90-minute executive committee readout.
What this is not
- ×We do not build, host, or resell tools. Implementation goes to your team or a chosen build partner identified in the rollout plan.
- ×We do not write production code in this engagement.
- ×We do not staff the build for you. The operating model defines roles; you fill them or we help you scope a build-partner search.
- ×We do not pitch a follow-on retainer at the readout. If you want ongoing partnership, that is the separate advisory retainer offering.
Pricing
$25,000 – $50,000
$25,000 covers an 8-week scope for a single-LOB firm under $5B in assets with a focused AI strategy (one or two priority workflows).
$50,000 covers a 16-week scope for a multi-LOB firm with regulatory complexity, multiple priority workflows, and board-level governance design.
Fixed fee. Paid 30% on engagement, 30% at midpoint phase-gate, 40% on delivery. Phase-gates have explicit sign-off and a kill option for both sides.
Recent representative work
Engagements anonymized for confidentiality. Marked cases are illustrative composites; unmarked cases are live engagements.
Asset manager
$22B AUM institutional managerIllustrativeSituation: Firm had run its own internal AI discovery and committed to investing $1.2M in AI capability across one fiscal year, but the program had no operating model. Three months in, the priority workflow had not started; everyone was waiting for someone else to own it.
Delivered: Designed the operating model with a named AI lead reporting to the COO, a governance committee with quarterly board reporting, and a 12-month rollout sequenced by phase-gate. Specified the data integration required between portfolio analytics and client reporting systems.
Outcome: Program launched within 4 weeks of the readout. RFP responses (the priority workflow) compressed from 5 weeks to 8 days within the first 90 days. Board's risk committee approved the governance frame on first review.
Regional bank
$18B in assets, multi-stateIllustrativeSituation: Bank had an assessment and a build appetite but did not have an operating model that the CRO and the CCO would sign off on. SR 11-7 model risk requirements were the blocker.
Delivered: Designed the model risk management addendum for AI specifically, with second-line review integration sized to the existing model risk framework. Drafted the governance committee charter. Sequenced three priority workflows.
Outcome: Bank's first priority workflow (CCAR data prep) went into production 5 months after the strategy engagement closed. The CRO publicly cited the engagement as the reason the program did not stall in committee review.
Common questions
Do we need to have done the assessment first?
No. Some firms come to us with their own internal discovery already done. We can start at strategy if the firm has a defensible view of where AI should pay back. If the discovery work is shaky, we will recommend running the assessment first.
Who from our side needs to participate?
An executive sponsor (usually the COO, CFO, or CEO), a project owner who is in working sessions weekly, and a stakeholder group of 6-10 people who participate in interviews and reviews. Calendar time for the sponsor is roughly 2 hours per week across the engagement.
What happens at a phase-gate kill?
Either side can call a kill at the midpoint phase-gate. If we call it, we refund the 40% delivery payment and explain the reason in writing. If you call it, you pay through the midpoint and walk with the work product to that point. Phase-gate kills are rare but they exist by design.
How is this different from a McKinsey or BCG strategy project?
Three differences. Fixed fee under $50K, not $500K+. Boutique team with named operators, not a pyramid with a partner on top. The deliverable is an operating model and a build plan with kill conditions, not a 200-page deck.
Need an operating model around your AI strategy?
Send a one-page brief to pat@clearhaven.ai or book a 30-minute discovery call. We will respond within one business day with a fit assessment.