The requirements document fixes the use case, the guardrails, the KPIs, and what success looks like. The first 90 days cover monitoring, onboarding, and refinements that stay inside that document.
AI consulting
AI consultingThe plan before anyone builds the AI.
Teams can see the potential of AI and still not know where to start, what to do first, or how to adopt it without disrupting the work. We write that plan behind your brand: a practical blueprint, then support so people actually use what gets launched. You keep the client.
What the consultancy is for
This is the strategic blueprint in front of a technical delivery. It is for digital agencies and the smaller businesses they serve. The aim is a plan that is practical, tied to a goal, and usable by the teams who have to live with it.
- 01
Clear use cases
Business objectives become AI and automation workflows someone can act on, including where time, cost, or revenue is the point.
- 02
Less scope creep
Ambiguity is reduced before a build, so the project is less likely to be reworked or delayed. Guardrails state what the AI should and should not automate.
- 03
Operational readiness
We look at whether the infrastructure and the data can support the AI, and at process maturity, current tools, and how comfortable people are with it.
- 04
Adoption, not a tool nobody opens
Training and change support are part of the work, so internal teams and end clients can use the new tools instead of leaving them on the shelf.
Features
Ongoing consultancy, after the blueprint
AI is not set and forget. After the first launch, three structures keep the work moving. None of them has a published price. The quote names which one applies.
Monthly
A monthly extension looks for new AI use cases, reviews performance and adoption, keeps tuning workflows and prompts, and supports governance and AI practice.
Quarterly
A quarterly addition is a higher-level look: AI performance and trends, plus scaling and maturity planning when a rollout needs to go wider.
Ad hoc
Short advisory, a second opinion on an existing third-party AI setup, a roadmap or prioritisation workshop, or scoping for a use case that does not fit the standard path.
Two phases
Discovery first, then the first 90 days
The service runs in two phases, before a line of code is written or an agent is built, and then while people learn to use what launched. You always know which stage you are in.
What the blueprint holds, and what is scoped again
A non-standard use case, a second opinion on someone else’s AI setup, or work that was not in the requirements document is scoped before it starts. Building the agent or the application is a separate project.
What we do, and what we need from you
Discovery stalls when the workflows, the objectives, or the people who will use the system stay unnamed. This is the split.
| Task | Your agency |  |
|---|---|---|
| Assess process maturity, current tools, and AI literacy, and name the opportunities worth doing first. | ||
| Define the use case against the business objectives. | ||
| Map data sources, the tech stack, and how people will interact with the system. | ||
| Set guardrails for what should and should not be automated, including compliance and technical constraints. | ||
| Deliver the requirements document with KPIs and acceptance criteria. | ||
| For the first 90 days, monitor behaviour, support adoption, and refine prompts, logic, and workflows from real use. | ||
| On a retainer, review performance and adoption, tune prompts, and identify the next workflow or department. | ||
| The current workflows, software, and data, so we can see where the business stands. | ||
| The business objectives and what the AI is meant to achieve. | ||
| Access to the people who will use it, so adoption is not designed in the abstract. | ||
| Feedback in the first 90 days, when edge cases show up in real work. | ||
| A decision on the requirements document before a build starts. | ||
| On a retainer, the priority for the next workflow or department. |
How agencies usually offer this
Use it to audit a client’s current workflows and find the manual bottlenecks AI can take. Use discovery before an AI build so the technical project is less likely to wander and the margin is protected.
Monthly and quarterly extensions can be packaged as retainers for ongoing optimisation. The same service can be used on your own internal workflows.
Why the strategy comes before the tool
Buying or building an AI tool without a strategy is the consultancy’s own comparison: a house without blueprints. Without a plan, the usual results are wasted budget, tools teams refuse to use, and missed opportunities.
The blueprint is there so the spend has a defined outcome: where AI can save time, cut cost, or raise revenue in this business, and so employees meet the change with the friction taken out.
The plan is allowed to change with the technology
Frameworks and optimisation steps are not frozen. They are updated as AI technology, platform capabilities, and compliance requirements move.
That update does not add a new use case by itself. A use case outside the requirements document is scoped before anyone starts it.
Agencies often add these when the blueprint is done
From the blog
Notes that sit next to this work. All notes
AI Solutions
How to Build an AI Marketing Strategy for Your Clients
A step-by-step approach to building an AI marketing strategy for clients, from auditing workflows to choosing use cases and measuring impact.
3 min read
AI Solutions
The Skills Marketing Teams Need to Work Well with AI
The practical skills marketers need to get real value from AI tools: briefing, critical review, data literacy, ethics and workflow design.
3 min read
AI Solutions
Using AI Responsibly in Agency Work: A Practical Framework
A practical framework for agencies using AI on client work: transparency, data protection, bias checks and human review without slowing delivery.
3 min read
Frequently asked questions
Agencies that want to offer AI consultancy under their own name, for their own workflows or for a client. The service is aimed at digital agencies and smaller businesses.
Ready to take the guesswork out of an AI project?
You keep the client. We look at the current processes and name the quickest wins before anyone builds.