AI TRAINING
Building the AI Center of Excellence
Leave with a funded, staffed CoE blueprint your board will approve and your teams will actually use.
What it covers
This programme equips transformation leaders and heads of data with the strategic and operational frameworks needed to design, launch, and sustain an AI Center of Excellence. Participants explore hub-and-spoke versus federated governance models, build a hiring roadmap, draft a CoE charter, and define an operating rhythm tied to business outcomes. Sessions combine case-study analysis, working-group exercises, and peer critique to produce artefacts participants can present to leadership immediately after completion.
What you'll be able to do
- Design and justify a CoE operating model (hub-and-spoke, federated, or hybrid) suited to your organisation's structure and AI maturity
- Draft a board-ready CoE charter defining mandate, scope, governance structure, and success metrics
- Build a phased hiring plan with role priorities, sourcing strategies, and retention levers for AI talent
- Define a funding model and internal chargeback mechanism that keeps the CoE financially sustainable beyond year one
- Establish an operating rhythm—including OKRs, steering committee cadence, and escalation paths—that embeds AI delivery into business-as-usual
Topics covered
- Hub-and-spoke vs federated vs centralised CoE operating models
- Writing and ratifying a CoE charter and mandate
- Hiring patterns: roles, sequencing, and build-vs-buy decisions
- Funding models: cost centre, profit centre, and internal chargeback
- Governance frameworks: AI ethics board, review gates, and policy layers
- Operating rhythm: OKRs, sprint cadences, and executive reporting
- Stakeholder alignment and change management across business units
- Measuring CoE ROI and demonstrating value to the board
Delivery
Typically delivered as a blended programme over 4-6 weeks: two in-person or virtual intensive days (kick-off and finale) bookending three to four facilitated online working sessions of 3-4 hours each. Participants work in cross-functional cohorts to build their own CoE blueprint in parallel with instruction. Materials include a CoE charter template, a RACI model for AI governance, a hiring-matrix workbook, and a funding-model calculator. Hands-on working time accounts for approximately 60% of total hours.
What makes it work
- Securing a named C-suite sponsor with budget authority before the CoE charter is written
- Embedding CoE liaisons inside high-priority business units from day one to maintain trust and adoption
- Publishing a transparent intake and prioritisation process so business units see the CoE as a partner, not a gatekeeper
- Running quarterly value reviews tied to board-level OKRs to sustain funding and political capital
Common mistakes
- Centralising all AI work in the CoE and creating a bottleneck rather than enabling distributed capability
- Launching without a clear mandate or executive sponsor, leaving the CoE to fight for budget every quarter
- Hiring senior data scientists first instead of engineering and MLOps roles that unblock delivery
- Measuring the CoE on project output rather than on business value delivered to internal customers
When NOT to take this
If an organisation has fewer than 50 employees or has not yet shipped a single AI proof-of-concept, a formal CoE structure will consume resources and create bureaucracy before there is enough AI work to justify it — a small AI task force or a single senior hire is a better fit at that stage.
Providers to consider
- DataScientestdatascientest.com →
- Wavestonewww.wavestone.com →
- McKinsey & Company (QuantumBlack Academy)www.mckinsey.com/capabilities/quantumblack/how-we-help-clients/capability-building →
- MIT Sloan Executive Educationexecutive.mit.edu/course/artificial-intelligence-implications-for-business-strategy/ →
Sources
This training is part of a Data & AI catalog built for leaders serious about execution. Take the free diagnostic to see which trainings your team needs.