Evaluate — step one
Find out what’s actually wrong.
Most engineering organizations know delivery is slower than it should be. Far fewer know why — because the visible symptom is rarely the cause. A Jump Start tells you, with evidence, ordered by what it is costing the goals you are being measured on — and hands you the finding whether or not you engage us to act on it.
- Scope
- Agreed in a first conversation
- Deliverable
- Written assessment, ranked to your goals
- Commitment after
- None
What we look at
Six things, in order of what gates the next one.
Which level you are actually on
Where you genuinely sit on the five levels of AI capability, against where your roadmap assumes you are. Almost always one or two levels apart, and the gap is the finding.
The data underneath it
Whether you could produce a clean, current dataset for a core process without a special project. Nothing above level 1 works if the answer is no, and this is the layer nobody budgets for.
Talent and structure
Who you have, what they are able to do, how they are leveled, and whether the org chart matches the way work actually moves — including who would own a model in production.
How work actually flows
The real path from idea to shipped, versus the documented one. The gap between the two is usually where the delay lives, and it caps how fast you can climb.
The toolchain
What you are running, what you are paying for, what nobody owns. A fragmented estate holds you at level 1 because the data cannot be joined.
Portfolio strategy
Where the money is going, whether it is going to the work you said mattered, and whether AI spend is assessed on outcomes rather than activity.
We talk to the people doing the work, not only the people reporting on it. Where the data exists we read it. Where it does not, that is itself a finding.
Included
AI readiness
We place you on the five-level capability ladder and say plainly where you actually are — which is usually one or two levels below where the roadmap assumes. Then we name which level your business goals actually require, which is not always the next one up, and what has to be true to get there: data quality, ownership, delivery cadence, governance, and the organization’s genuine capacity to absorb change. You get a current-state reading you can take to a board, whether or not you engage us to close the gap.
See the five levelsThe stages
What actually happens.
Five stages, and what comes out of them is ordered by something you set rather than by us. We agree the business goals at the start, and everything after that is ranked against them.
Scoping
The initial conversation: the business goals the work has to serve, what you want answered, which teams and sites are in scope, and what access we need. The goals are written down here because they are what the findings get ranked against later. This is also where the length is agreed, and it is agreed before you commit to anything.
Interviews
We talk to the people doing the work, not only the people reporting on it. Engineers, leads, product, and whoever is closest to the part that hurts.
Evidence
System and toolchain review, portfolio and delivery data. Where the data exists we read it; where it does not, that is itself a finding.
Analysis
Constraints ordered by what they cost the business goals agreed in scoping, not by ease of fixing. Two organizations with the same constraints get different orders, because they are not chasing the same thing. The ranking is the part that takes judgement, and it is done by the operators who ran the interviews.
Delivery
The written assessment, delivered and walked through with your leadership team. Yours to keep, act on, or ignore.
How long it takes
That depends on you, and we work it out together in an initial conversation — how many teams, how many sites, how much of the evidence already exists. We scope it before you commit to it, and then we hold to what we scoped. Quoting a fixed duration before looking would be the same mistake as quoting a fixed remedy before looking, and we say elsewhere on this page that we do not do that.
What you get
A written assessment naming the specific constraints on delivery, ranked by what each one is costing the business goals you set out at the start, with a recommended sequence for addressing them. Yours to keep and act on, with or without us.
Who runs it
Operators who have held the roles they are assessing — not analysts, and not a junior team with a template. The same people who would drive the work if you asked us to do it.
What decides the order
The business goals you name at the start, written down before anyone looks. We rank the constraints by what each one is costing the things you actually have to deliver — the launch you have committed to, the cost you have to take out, the audit you have to pass — not by which is easiest for us to fix or quickest to show progress on. If the goals change during the engagement, the order changes with them.
Start with a two-hour discovery.
Two hours with the people who would run the work. We listen, we tell you whether we recognise the problem, and we say what we think it would take. No obligation to proceed, and no deck.