Enterprise AI Delivery Assessment

Turn an AI opportunity into a secure, practical delivery plan.

We assess your candidate use cases, identify the strongest starting point and define the architecture, security controls, governance and acceptance criteria needed to move forward.

Book a qualification call

A paid, fixed-scope engagement. We agree your deliverables and fixed fee before work begins.

Two weeks · Up to three candidate use cases · One detailed delivery recommendation

Who this is for

You have a delivery problem worth solving and need to establish where AI can help, what it can safely access and how you will judge the result. The assessment gives a CIO, CTO or transformation sponsor a practical starting point, supported by the people who will own delivery and operations.

• Software delivery: a stalled initiative, internal tool or backlog that needs a clear scope and acceptance criteria.
• Migration: an existing system that is difficult to maintain, with behaviour, data or public interfaces that must survive the change.
• Technical knowledge: specialist documents and experience that people need to find, understand and apply during their work.

Bring a business problem and a sponsor. You do not need to arrive with a chosen model, agent framework or finished solution design.

What you receive

A prioritised opportunity shortlist

We compare up to three candidate use cases against their business value, suitability for AI, available evidence, dependencies and delivery difficulty. You receive a recommendation for the strongest starting point, with the reasons behind it.

Current-state findings and risks

For the preferred candidate, we examine the relevant systems, source material, data constraints and delivery process. The findings identify what is ready, what needs to change and which assumptions still need checking.

A secure delivery blueprint

We define one bounded pilot: its scope, proposed architecture, data handling, agent permissions, review points and acceptance criteria. The blueprint explains what can be delegated to AI, where people must decide and what evidence will demonstrate an acceptable result. It also covers rollout, handover and operational ownership.

An executive decision brief

You receive a proceed, reshape or defer recommendation, supported by an indicative effort and cost range, assumptions, milestones and next steps. We walk through it with your sponsor and delivery owner so the next decision is explicit.

The assessment estimates a potential pilot; it does not commit you to implementation or guarantee a saving.

How the two weeks work

The two-week delivery window starts once the agreed information, access and stakeholders are available. We confirm these prerequisites before scheduling the engagement.

Stage	What we work through	What you take away
Discover	Business problem and up to three candidate use cases	A prioritised shortlist with value, feasibility and constraints
Assess	The preferred candidate’s systems, data, delivery process and security requirements	Findings, dependencies and risks
Design	One bounded pilot, its architecture, controls and verification	A delivery blueprint with acceptance criteria
Decide	The recommendation and investment required	An executive readout and pilot plan with indicative effort and cost

Your team provides a sponsor, a delivery owner, relevant engineering and security representatives, and the agreed documentation or source material. We agree participation during qualification so the work has access to the people who can resolve key decisions.

Security throughout

Security decisions belong in the delivery plan from the start. The assessment makes five areas explicit:

• Data handling: what information is involved, where it may be processed and what restrictions apply.
• Agent access and permissions: which tools and environments an agent needs, with access limited to its agreed task.
• Human approval points: which actions require review or authorisation, and who owns those decisions.
• Auditability: what records are needed to understand actions, review changes and investigate a problem.
• Operational ownership: who maintains the result, manages access and responds when something goes wrong.

Production access is considered only where necessary and explicitly agreed. Any unresolved security or governance decisions remain visible in the recommendation.

Delivery in practice

These projects show the delivery approach and its checks. The assessment establishes how that approach fits your organisation’s systems, risks and intended outcome.

An internal website migration

Our own website moved from a WordPress export to structured, AI-editable source. The work included content comparisons, release checks and recovery of legacy media used outside the site. It demonstrates why migration acceptance criteria must cover the system’s existing users and interfaces.

Read the website migration case study.

A personal technical knowledge project

The Otari resource brings scanned manuals, navigable digital editions, source comparisons and interactive explanations together. Its owner used that resource and AI-assisted research to complete a recap and calibration. The software can be inspected publicly; the physical outcome is owner-reported.

Read the Otari case study.

Scope and commercial terms

• Fixed fee after qualification. We confirm suitability, scope, deliverables and the fee before work begins.
• Up to three candidates, one detailed recommendation. The assessment develops one preferred initiative into a delivery blueprint.
• One consolidated feedback round. Your team brings its comments together for review against the agreed deliverables.
• Implementation is separate. A prototype or production build is not included as a standard assessment deliverable. Any feasibility exercise is explicitly agreed within scope.
• Specialist assurance is separate. Penetration testing and formal compliance certification require their own scope. This assessment does not certify compliance or guarantee security.

You can use the recommendation to decide your next step. There is no requirement to commission the implementation with us.

Questions before you start

What should we bring to the qualification call?

A short description of the problem, the intended outcome, the likely sponsor and the systems or source material involved. We use the call to establish fit and arrange the next steps. Do not send credentials, sensitive records or source-code uploads with the initial enquiry.

When does the two-week window start?

Once the agreed information, access and stakeholders are available. We confirm the start date and required participation together; the clock does not start when you submit an enquiry.

What access do you need?

That depends on the candidate initiative. We agree the documentation, source material and access needed for the assessment during scoping. Production access is considered only where necessary and explicitly agreed.

Will we get a working prototype?

The standard deliverable is an assessment and delivery plan. If a focused feasibility check is necessary to answer a specific question, we agree its scope first. A working prototype or broader implementation is a separately scoped engagement.

What if the recommendation is to defer?

You receive the reasons, dependencies and conditions that would make the initiative worth revisiting. Reshaping the scope or choosing a smaller improvement can also be the right result. The assessment supports an informed decision, including a decision not to proceed.

Start with a qualification call

Bring the problem you want to solve. We will discuss whether the assessment fits, the people and information it would need, and what a useful delivery decision looks like for your organisation.