AI Readiness & Process Assessment

• Process and service inventory. Every candidate captured at the level of the individual step, not the department.
• AI suitability scoring. Each process scored for readiness, complexity and volume, so the ranking is comparable across very different work.
• Data and document readiness. Where the inputs actually live, what condition they are in, and what is missing.
• Systems and integration reality. Which platforms expose an interface, which do not, and what that costs.
• Governance and sensitivity. What classification the data carries and which decisions require a human by policy.
• Operating model. Who owns the outcome after launch, and whether that role exists today.

• A scored inventory of every process and service, ranked by impact against effort.
• A phased delivery plan aligned to your own launch calendar, with the reasoning recorded against each decision.
• A written rationale per item, including why work was excluded. Exclusions are as valuable as inclusions.
• An executive scorecard suitable for a board or steering committee.
• A first-90-days plan with decision gates that separate a funded pilot from an experiment.

Most assessments return a maturity score and a slide pack. We return a sequence you can fund, because every score carries the reasoning behind it and every exclusion is explained.

We also tell you what not to do. On a recent federal engagement the assessment concluded that roughly a third of the operational processes reviewed were unsuitable for AI, because they run inside platforms the entity does not control. Naming that early protected both the budget and the timeline. An assessment that finds everything suitable has not been done properly.

How long does an AI readiness assessment take?
Three to six weeks for a department, depending on how many processes are in scope and how available the subject-matter experts are. A full entity-wide assessment covering operations and citizen services typically runs six to ten weeks. If a supplier quotes a quarter, they are usually rebuilding your process documentation rather than assessing it.

What makes a process a good candidate for AI?
Four things together: the rules are written down somewhere, the data is structured enough to act on, the volume justifies the build, and a human is willing to be accountable for the outcome. A process missing any one of those is better automated conventionally, or left alone.

Do we need clean data before we start?
No, but you need to know how dirty it is. Part of the assessment is establishing the actual condition of your data and documents, because that determines whether a process is a three-week build or a three-month one. Discovering it afterwards is what causes overruns.