Skip to content

CUSCOE operating model

CUSCOE is an operating model for organisations that combine professional services and managed services and are moving to deliver AI as a service. It was created by Damian Copeland and is owned by CX Bridge Ltd. It is published under CC BY-NC-SA 4.0, with an additional permission for internal use set out in Permissions.

The model puts a single governed Portfolio at the centre of the business. Technical people are pooled into competency-based Centres of Excellence, Customer Success owns the customer from the moment the contract is signed, and every other function connects to the Portfolio rather than to individual deals.

Why the model exists

AI systems are not finished at go-live. Language models are non-deterministic, and the data they work on changes over time, so a production AI service needs permanent monitoring, evaluation and tuning. That work needs the skill level that has traditionally sat in project teams, not in support.

This breaks the usual structure of a services business. Pre-sales, professional services and managed services are normally separate vertical teams working at different skill levels, with escalation running upwards from support to engineering. Once the skilled work moves downstream, the organisation needs to put the right person on the right task quickly, whatever team they sit in.

Commercial structures break too. When sales owns the customer after signature, there is a steady pull towards selling things the business cannot deliver or support. Fixed-price contracts do not survive variable inference costs and unpredictable end-user behaviour. Finance, legal, compliance, IT and the board all have to change how they work, and that change does not happen without someone at the top owning it.

How the model fits together

  • Portfolio: the constitution of the business. Products, pricing, delivery standards and legal clauses, defined once.
  • Centres of Excellence: where the value is made. Competency-based pools of technical people serving pre-sales, projects and ongoing operation.
  • Customer Success: owns the customer relationship and the account P&L from signature. The one commercial route to the customer.
  • Sales: wins new customers and sells only from the Portfolio.
  • Pricing model: the three-layer structure for selling services that contain AI.
  • Finance: approves the cost model once at Portfolio level and re-engages only on outliers.
  • Legal and compliance: defines clauses, acceptable use and AI governance once, at Portfolio level.
  • IT: owns infrastructure and security, and reviews the tools the Centres of Excellence select.
  • Board: sits outside the day-to-day model and holds named accountability for AI.

The design principles page lists all eleven principles in one place, with the evidence behind each.

The two rules

Two rules hold the model together. If either is broken routinely, the model stops working.

  • No sale outside the Portfolio without written approval. Anything sold, delivered or committed has to come from the Portfolio, and changing the Portfolio takes written approval.
  • No commercial contact between a Centre of Excellence and the customer. Centre of Excellence members deliver directly to customers, but scope, pricing and delivery-standard disputes always route through Customer Success.

Evidence status

The model is published as it stands, with each part labelled by how far it has been tested. All eleven principles have been evaluated against an illustrative scenario. None has yet been tested in a published live study. See Design principles for the detail and About for how to cite the model.