Why, who, where
and how AI should
operate.
This is how our AI engagements work. We call it operating architecture: settling why, who, where and how AI operates before anything gets built. Your leadership team settles all four, with us working alongside them as Architect, in a sprint or through guidance. Builders then make the approved solutions real. GEM Blueprint governs, enables and measures the whole lifecycle, from the first decision through to what it's actually delivering.
Built for CEOs and boards of technology driven businesses who already know spend isn't the constraint, clarity is.
AI is already operating in your business. Nobody decided it should.
Most organisations are running this in the wrong order. Tools first. Pilots first. Vendors first. AI is already inside the business, in someone's inbox, someone's spreadsheet, someone's customer conversation, long before anyone with the authority to decide has decided it should be. Then governance gets scrambled together after the fact, when a board member, a customer or a regulator asks a question nobody in the room can answer with confidence.
Technology is not the hard part. Deciding where it belongs is.
Long before AI, the businesses that struggled with technology were rarely short of tools or budget. They were short of leadership that knew where technology belonged and could hold the line on it. That is the technology leadership work our Managing Partner has done for twenty five years, and it is still what decides whether AI creates value or just noise.
Deciding where AI belongs is the same leadership problem with higher stakes. That decision has to happen before any tool gets selected, not after, and it has to be made by the people who carry the consequences: the CEO, the board, and the leadership team. That is why the work starts with why, who, where and how, settled with your leadership team in the room, not with a product demo.
GEM Blueprint
One framework surrounds the entire lifecycle, from the first decision through to what a use case is actually delivering.
Four questions. Four working artefacts.
One decision each.
Each question ends in one artefact your leadership team owns. Together they are the brief a Builder works from.
Why
AI Purpose Statement
Why AI matters to this business and the outcomes it must support. Without this, every subsequent decision has nothing to be measured against, and teams end up justifying tools after the fact instead of before it.
Who
AI Champion and accountable owners
Who leads, who owns, who reviews and who escalates. Most organisations we meet can't answer this cleanly. AI activity is happening, but no single person is accountable for the direction of it.
Where
AI Operating Map
Where AI is operating today, where it could operate, and where it should. A live record, not a one-off audit. The Map is what makes the board's question ("where is AI in our business?") answerable on demand.
How
Solution definitions
What an approved solution must achieve, its boundaries, its oversight and how it's measured. This is the brief a Builder works from, specific enough to build against, not a vague aspiration.
Governance, enablement and measurement. Across the whole lifecycle, not just at the start.
Governance
Sets the boundaries and the accountability. Who can approve what, what needs sign-off, what the escalation path is when something goes wrong. Established once, referenced constantly.
Enablement
Creates the capability and the adoption. A policy nobody understands and no one can use doesn't change behaviour. Enablement is what turns a document into how the business actually works.
Measurement
Determines whether AI is actually producing value, and whether a use case should continue, change, pause or stop. Without this, "we implemented AI" and "AI improved the work" stay two different, unrelated claims.
Architects and Builders.
Two roles, kept deliberately separate. Techshin is your Architect partner: the decisions stay with your leadership team, we bring the discipline. We help you find and brief the right Builder when one is needed. We are never the ones building the solution ourselves.
| Role | Responsibility | Typical outputs |
|---|---|---|
| Architect | Purpose, ownership, mapping, prioritisation, governance and solution definition. | Purpose Statement, AI Champion, AI Operating Map, guardrails, decision register, solution brief and measures. |
| Builder | Prototype, technical design, integration, deployment and maintenance. | Prototype, workflow, automation, agent, integration, deployment record and operating support. |
Where a Builder is required, Techshin helps prepare the brief, evaluate capability and review delivery against the approved purpose, guardrails and measures, and stays on the leadership side of that decision the whole way through.
Frame. Activate. Advance.
The order the work actually happens in. Frame is where the four questions get settled. Activate is where something gets built. Advance is where leadership decides what happens next, and the cycle starts again.
Frame
Decide what good looks like and what role AI should play, before anything is built. This is the Architect's work, done with your leadership team.
Why
Purpose before tools. What AI is actually for in this business, agreed at leadership level.
Who
An accountable owner named, with real authority to lead, review and escalate.
Where
Visibility across the business through the AI Operating Map, a live record, kept current.
How
Solution definitions written for anything approved to move forward, specific enough for a Builder to work from.
Activate
Take the decision made in Frame and get it operating for real, at whatever autonomy level was chosen.
Augment
A person does the work with AI beside them. Often as simple as a tool and a well written prompt, with a human reviewing every output before it matters.
Automate
A dedicated tool or workflow runs a defined process end to end. No judgement asked of the AI, clear inputs and outputs, a named owner watching it.
Agentic
AI acts with more autonomy inside guardrails. Rarely the right call where judgement or oversight sit close to the work. Frame decides which of the three applies.
Whichever level, an internal team or a Builder builds it against the solution definition, and enablement makes sure people can actually use it.
Advance
Measure what is operating and decide what is next: more autonomy, less, a replacement, or stopping it entirely.
Measure
What it is actually delivering against the business case it was approved on. Value, risk and adoption, through GEM Blueprint.
Decide
More autonomy, less autonomy, a replacement, or stop it entirely. Made by the named owner, not left to drift.
Update the Map
The decision goes back onto the AI Operating Map, ownership is confirmed, and that piece of work reopens Frame.
Then back to Frame. You cycle it, you don't graduate from it. Every Advance decision reopens Frame for that piece of work, which is why governance, enablement and measurement have to run across the whole cycle, not just at the start.
Questions leadership teams actually ask.
Why not just start with a tool or a pilot?
You can, and plenty of organisations do. The exposure shows up later, when the board asks where AI is operating and nobody can answer, or when three teams have quietly built three different answers to the same problem. Frame first, and that question stays answerable.
Do you build the AI yourselves?
No. Techshin is your Architect partner. Your leadership team settles why, who, where and how, and what an approved solution must achieve, with us working alongside them in a sprint or through guidance. Builders make it real, either your own internal team or a Builder we help you brief and select.
How long does this take?
The AI Framing Sprint establishes Why, Who, Where and How in six weeks, with your leadership team in the room. It's a fixed engagement, not an open-ended consulting retainer.
What happens after the framing is done?
The AI Operating Map stays live as the system of record, and Map Advisory keeps leadership judgement current as new opportunities and decisions come up: reviewing the Map, challenging proposed solutions, and deciding what proceeds, pauses or stops.
What's actually different about GEM Blueprint versus a normal AI governance framework?
Most governance frameworks stop at the boundary-setting. GEM Blueprint carries Governance, Enablement and Measurement across the whole lifecycle. It's still doing work after a solution is live, not just before one is approved.
See it in a six week engagement.
The AI Framing Sprint is how Why, Who, Where and How get established with your leadership team, in the room, in six weeks.