Vaughn Harvey & Co

Field-tested, stripped of client detail

Where you are decides where we start.

Every playbook is scoped with named outcomes, monthly deliverables and tollgates, and a defined end.

The 90-day AI transformation program

One program, three workstreams, running in parallel from day one.

Month 1

Organized, prioritized, baselined

Transformation Office
Office objectives; target-state org model and the permanent owner role; year-one run rate; build, hire or contract recommendation; facilitated prioritization into one 90-day plan
People Agenda
Rules of engagement v1; role-by-role AI proficiency matrix for priority roles; champions selection criteria and cohort
AI Governance
Governance gap assessment and priority risks; current technology landscape and architecture review; AI risk mapped to cyber and D&O coverage

Month 2

Running

Transformation Office
Workflow inventory ranked for efficiency and growth; first workflow selected; owner role defined
People Agenda
Proficiency matrix complete; champions program launched with method and guardrails
AI Governance
Design blueprint and recommendations to close the gaps: guardrails, review gates, governed self-serve, skills and connector intake

Month 3

Results and renewal

Transformation Office
First workflow reimagined as a design: agentic target state, roadmap and a business case Finance can use; adoption and cost scorecard
People Agenda
Training curriculum v1; certification design; champions operating rhythm
AI Governance
Governance scorecard; model and cost governance design; execution plan

Field-tested now at a growth-stage fintech: a four-person team, with Month 1 delivered on schedule.

Where this fits

  1. Program
  2. Build
  3. Install
  4. Maintain

Playbook 1 · Transformation Office: reimagine a workflow

Six steps from charter to a fundable design.
  1. 01

    Select and scope.

    One-page charter: boundaries, stakeholders, the metrics Finance will judge it on.

  2. 02

    Map the current state.

    How the work actually flows: hand-offs, systems, data, volumes, cycle times, controls.

  3. 03

    Assess AI and agent suitability.

    Task by task: automate, augment or eliminate; human-in-the-loop and control points.

  4. 04

    Design the agentic target state.

    Which steps agents run, how they are orchestrated, the human roles that remain, the guardrails.

  5. 05

    Build the business case.

    Cycle time, cost to serve, quality, capacity released; cost to build and run; payback.

  6. 06

    Sequence the roadmap.

    Build order, a scoped pilot, data and technical prerequisites, adoption plan and risks.

In wealth management the first workflows are usually client onboarding, account servicing, the supervision and surveillance queue, and the advisor's meeting-prep-to-follow-up loop.

Playbook 2 · People Agenda: champions, proficiency by role, rules that stick

  1. 01

    Rules of engagement.

    What people may and may not do with AI, in plain language, published once the owner is set and refreshed as the guardrails mature.

  2. 02

    AI proficiency matrix.

    Role by role, what good looks like: for an advisor, an operations lead, a compliance officer, a partner. The matrix drives training, hiring and the champion criteria.

  3. 03

    AI champions program.

    Selection criteria, a confirmed cohort, an operating rhythm, and re-seating on usage data rather than enthusiasm.

  4. 04

    Learning path and adoption.

    Curriculum and a certification design the firm delivers itself: use cases, not 101. Adoption measured by role and line, so the learning path reaches the reluctant tail as well as the power users.

Playbook 3 · AI Governance

A practical governance framework covering 45 elements, drawing on ISO/IEC 42001, the EU AI Act and DCAM. Clear accountability and controls that let adoption move with confidence.
Governance target state.
The ownership, policies and decision structure the firm needs to govern AI across business and technology, aligned to its risk appetite.
Risk and control readiness.
A view of the material gaps across AI use, data, security and third-party capabilities, with priorities the leadership team can act on.
Operating model.
Responsibilities for the committee and control owners, and a proportionate approach to review, oversight and ongoing accountability.
Roadmap and assurance.
A sequenced improvement plan and a governance scorecard for leadership, grounded in the standards and regulatory obligations that apply to the firm.

The governance work establishes the direction and the controls; the AI governance platform supports their day-to-day operation.

Playbook 4 · AI Platform Strategy

For firms deciding what to run AI on. A target-state architecture and the decisions that get you there, framed vendor-neutral.
Target state.
The enterprise AI platform: model access and routing, orchestration, agent runtime, the knowledge layer, identity, observability and cost controls, drawn as one picture the committee can govern.
Knowledge layer.
The firm's structured and unstructured knowledge made AI-ready and owned in-house as a product, not held by a consultant.
Data foundation.
Platform and warehouse choice framed vendor-neutral; the data architect role designed and the search supported.
Partner strategy.
Which capabilities to buy, which to build, which vendors to run alongside, and the exit from each.
Taxonomy and measurement.
A workflow taxonomy that ties usage to outcomes, so the board sees workflows automated and value produced, not prompt counts.

Run now for two insurance and retirement businesses through a merger, where the two estates have to become one platform.

Talk through which playbook fits.

Book a working session