AI Workflow Assessment · two weeks · fixed fee
A two-week, fixed-fee assessment, quoted before you commit, for businesses that want AI working on real processes and need to know where it will pay. You get processes ranked by value, the authority each step needs, cost ranges and a clear 'not yet' list.
01What you are trying to do
The board is asking, vendors are pitching and customers expect more. What is missing is a clear view of which of your own processes AI can improve, and what that would be worth.
02What gets in the way
Both have a cost. A fixed-fee assessment of what can safely be automated, quoted before you commit, gives you a written plan the board can back before larger sums are spent.
03How we help
We will tell you what to automate, what to modernise first, what to wait on, and what not to do at all.
01 / How it works
You bring the people and the documents. We do the analysis. You leave with decisions you can act on, whether or not you continue with us.
02 / The shape of the work
The output is a short, defensible document that your board and your risk team can both act on.
We map the processes that cost your people the most time, the systems and data they touch, and the identity layer that already governs access. We name the constraints automation has to live within.
For each candidate process we separate the steps rules can handle, the steps that need AI judgement, and the steps that need a person. Then we map the authority each step should have.
A phased roadmap with cost ranges: what to automate now, what needs modernising first, what to do next quarter, and what to leave alone for now.
03 / What we need from you
The assessment is read-only and works from conversations and documents. It covers the processes and systems agreed in the quote, and the fee is set by how many are in scope.
The documentation you already have: process notes, policies, procedures and system or architecture diagrams. No source code or production access is needed.
Time with the people who own each process in scope and the people who do the work day to day, for walkthroughs and conversations. We plan the sessions with you.
The quote names the processes and systems in scope. The fee depends on how many there are, and work outside them is agreed separately.
Where private inference is an option, the comparison with commercial APIs is an estimate on a representative workload, with the assumptions stated. It is not a measured benchmark unless one is separately agreed.
04 / What you get
All six are designed to be useful even if you decide not to proceed with us.
Candidate processes ranked by value and by risk, based on how your business actually runs.
For each step: Assist, Prepare, Execute or Escalate, with who approves what. The foundation of the governance design.
Which systems automation can reach safely today, and which need modernising, wrapping or integrating first.
Where the AI should run for each class of data: private models, approved commercial models, or a hybrid. Where private inference is an option, we compare it with commercial APIs on a representative workload, counting infrastructure, implementation, support, retries and human review against the same quality and service requirements. It is an estimate with its assumptions stated, not a measured benchmark, unless a benchmark is separately agreed.
Time-boxed phases and budget ranges. A defensible answer to 'how much, how long, what for'.
The processes we recommend you do not automate yet, and why. Often the most useful page in the document.
Common questions
No. It is fixed-scope and fixed-fee, and it ends in a written deliverable whether or not you go on to work with us.
It is a fixed fee, quoted before you commit. The shape is two weeks of senior architect time plus a written deliverable. The exact number depends on how many processes and systems are in scope, and we are explicit about that on the first call.
It replaces it. An AI readiness assessment asks whether you are ready for AI. This one asks which of your processes AI should help run, under what authority, and what has to be true first. It is a more useful question.
That is a legitimate outcome and we will say so. The deliverable is still useful because it tells you what to revisit and when.
Some time from the people who own each process in scope and the people who do the work, for walkthroughs and conversations, plus access to the documentation you already have. We plan the sessions with you so they fit around the working week.
No. The assessment is read-only and works from documentation, architecture diagrams, process walkthroughs and conversations with your team. If we go on to deliver the work, that conversation is separate.
Two ways in
Not ready for two weeks? Start with a Fit call. A free 30-minute call with a senior engineer. Bring one problem. We tell you whether it is worth a longer conversation, and if not, who you should talk to instead.
