AI Readiness Assessment
AI Readiness Assessment
What your systems will support, which jobs are worth doing, and what gets built first.
What an assessment is
Most AI projects stall because nobody decided what was being built before the building started. An assessment produces that decision as a written document.
It answers three questions: what your systems will support today, which jobs are worth doing, and in what order they should be built.
It is deliberately separate from the build. The output is useful whether CloudCentric builds it, somebody else does, or you decide not to.
What you get
The short review, and the full assessment
Two different pieces of work. The short one is free and starts a conversation. The full one is a paid engagement and produces a document you could hand to any builder.
Short review · no charge
Where your documents live, whether the tenant will support the work, and the first job worth doing. Answered in writing, usually within a week of a conversation.
Full assessment · paid engagement
Every candidate use case found and ranked, the tenant and data findings written up, the architecture chosen and drawn, and a build order with the dependencies marked.
Use case inventory
The jobs people actually do that a system could do instead, each with what it currently costs in time and how hard it would be.
Tenant and identity findings
Licensing, identity, network position and the permission state of the content the system would read.
The architecture, drawn
Which platform, which deployment type, where retrieval sits, and how it reaches the systems of record.
A build order
What gets built first, what it depends on, and what should wait.
How it works
How it runs
A conversation
What the business does, where the repetitive work is, and what has already been tried.
Look at the tenant
Licensing, identity, where the content lives and what state its permissions are in.
Interview the people doing the work
The candidate jobs come from the people who currently do them, not from a list of industry use cases.
Write it up
The findings, the architecture and the build order, as a document that stands on its own.
In practice
What comes out of it
The document is the deliverable. This is what is in it.
A ranked list of candidates
Every job a system could do instead of a person, with what it currently costs in hours and how hard it would be to build. Ranked, so the argument about what to do first is already settled.
The findings on your tenant
Licensing you already hold, identity configuration, where content actually lives, and the permission state of the material a system would read.
The architecture, drawn
Which platform, which deployment type, where retrieval sits and how it reaches your systems of record — as a diagram, not a paragraph.
A build order with dependencies
What is built first, what has to exist before each piece, and what is better left alone. Including the jobs that should not be built.
What matters
What the assessment usually finds
Recurring findings, across the kinds of business that ask for this.
The best first job is rarely the one they came in asking about
The job people name is usually the most visible one. The one worth doing first is usually more boring and much easier to finish.
The documents are not where people think
Content is normally spread across SharePoint, a file server, somebody's mailbox and a filing cabinet. Finding it is part of the work.
Permissions have drifted
Content that accumulated over a decade usually has broader access than anybody intends. That gets reviewed before an assistant can read it.
Licences already held cover more than expected
Most firms on Microsoft 365 already own a meaningful part of what they need and have not switched it on.
One document type is doing most of the damage
There is normally a single recurring document that costs more staff hours than everything else put together.
Some of it should not be built
Part of the value is naming the jobs where the effort would not return and saying so plainly.
Who it is for
Who asks for one
Firms who have been asked by their board
Somebody senior wants a position on AI and there is no technical answer to give them yet.
Firms whose staff are already using it
People have started pasting work into consumer chat tools and the business wants a supported route instead.
Firms with one obvious bottleneck
A known pile of repetitive document work, and a question about whether it can be dealt with.
Firms who got a proposal they cannot judge
A quote arrived from somebody else and there is no way to tell whether it is sensible.
Questions
Frequently asked questions
What does the assessment cost?
The short review is free. The full assessment is quoted against the size of the estate and the number of use cases in scope — call and we will scope it.
How long does it take?
The short review is usually a week. A full assessment on a single-site business is normally two to three weeks.
Do we have to build with you afterwards?
No. The document is written so that any competent builder could work from it, and several clients have used it to brief someone else.
What do we get at the end?
A written document: the ranked use cases, the tenant findings, the architecture and the build order.
Is this the same as a Copilot readiness check?
It covers that, and more. A Copilot check asks whether one licence will work. This asks what should be built, including the work Copilot cannot do.
Who needs to be involved?
Whoever administers the tenant, and two or three of the people who do the repetitive work. It is not a workshop for the whole company.
Also on this site
Start with the short review
No charge, and it answers whether there is anything here worth building.
(844) 422-7000