Case · AI delivery

Naming the roles AI delivery needs before the market had the words.

The gap in AI is not strategy, it is delivery — and delivery needs roles the old playbook does not contain. These are the operating concepts defined to fill it, and what each one changes about who answers for the result.

Concepts
Four, defined and in use
Roles
Two, named and staffed
Span
2026 → now
Role
Originated the concepts and the role definitions
01 / The situation

Funded ambition, and nobody accountable for the outcome.

  • Pilots are run, demonstrated, and never reach delivery.
  • Tools are adopted and the way the work happens does not change.
  • Engineering is measured on tickets closed while the business waits for a result.
  • No role exists whose job is the outcome rather than the backlog.
02 / The concepts

Four ideas, each one moving accountability.

01

Outcome Engineering

Engineering held accountable for the result and the artifact, not for the closed ticket. It changes what "done" means, and therefore what gets measured.

02

Forward Deployed Engineers

Specialists placed inside the business who own the outcome rather than a backlog — close enough to the problem to change the solution, and answerable for whether it works.

03

AI Workflow Engineering

Designing the workflow itself instead of buying tools and hoping: what the machine does, what a person decides, and where the hand-off sits.

04

AI-augmented SDLC

AI embedded across the delivery lifecycle — with an honest account of where it earns its place and where it does not.

03 / Why naming matters

A role without a name cannot be hired, measured or defended.

Defining a role is not vocabulary. It is what makes the work budgetable.

  • A named role can be written into a job description and hired against.
  • It can be given a measure, and therefore reviewed.
  • It can be defended in a budget conversation, because it is a thing rather than an idea.
  • And it can be refused deliberately — which is its own kind of clarity.

Intelligence is cheap now. What remains scarce is somebody whose job is the outcome.

If AI is funded in your organization and still has not changed how delivery works, the missing piece is usually a role, not a model.

Scope a transformation