Skip to content

01

Data and AI readiness diagnostic

An inspection of the system you actually run, carried out by the person who will do the remediation if you decide to go ahead. It produces a written report you can act on, hand to another vendor, or file.

Commercial terms

One fixed fee for two to three weeks. What moves the number is how many systems feed the documents, not how many days we spend — so it is quoted once we have seen the inventory, and it does not change afterwards. It needs about six hours of your team's time across the engagement: access setup, two operator interviews and the review. Prerequisites are read-only access to pipeline definitions and job schedules, an environment we can restore into, and one person who runs the system day to day.

Not the right fit if

  • A team that wants a maturity score to present upward
  • An organisation looking for a document that says everything is fine
  • A greenfield build — there is nothing yet to inspect

The problem

Most data platforms are not failing loudly. They are failing quietly: a report that has drifted, a nightly job nobody owns, a tenant boundary that was assumed rather than tested.

Nobody notices until an auditor, a regulator or a customer asks a specific question — and the honest answer is that nobody is certain how the number was produced.

The diagnostic exists to find those answers before someone else does, while there is still time to fix them cheaply.

What we actually do

  • Read the pipeline inventory: what is scheduled, by what, owned by whom, and what happens when it fails
  • Examine tenant isolation directly — row-level policies, role separation, and whether a query can cross a boundary
  • Inspect credential handling and the blast radius of each service identity
  • Attempt a restore into a disposable instance and time it against the stated objective
  • Trace where model output reaches a decision, and what evidence accompanies it
  • Interview the people who actually operate the system, not only the people who bought it

What you get

What you get.

Everything produced is written down and handed over. Nothing lives only in someone's head, and nothing depends on us to keep running.

  • A written report ordered by risk, with a finding, an impact and a remediation for each item
  • The measured restore time, not an estimate
  • A remediation sequence with effort indicated, so it can be budgeted
  • A 60-minute review with your technical and risk stakeholders
  • An optional follow-up question-and-answer window after the review

Who it is for

A data platform that has grown by accretion and has no current owner of its architecture
A claims, billing or reimbursement process where a rounding difference becomes a dispute
An AI initiative a business unit wants and a risk function has stalled
A team preparing for due diligence, an audit, or a change of platform

Next step

A thirty-minute scoping call. Bring the problem rather than a specification.

Book a conversation