AI & Data Readiness Assessment

Who It’s For  

CEOs and leadership teams who know AI matters but aren’t sure where to begin. Usually the company has grown by adding software: a system for accounting, another for inventory, another for sales channels, and a few spreadsheets holding it all together. Each one does its job. None of them give leadership the whole picture.

The Problems It Solves  

• Your data is scattered. It sits across subscriptions, and each vendor controls its piece.
• Your controls run on manual work. Before anyone can review the numbers, someone spends days gathering them, rebuilding them in Excel, and tying them out.
• Everyone is asking about AI. Vendors are selling it, the board is asking about it, and nobody has a plan that starts with your own data.

How It Works  

1. Walk through the work with the people who do it. We map each key process, step by step, and review the map with the people who run it.
2. Read the actual files. Close books, templates, reconciliations, the controls matrix. The real work lives in these files, so that’s where we look.
3. Map the current state: every system, the data it holds, how data moves between systems or doesn’t, and what each subscription costs. We confirm it with finance, IT, and operations.
4. Assess the controls: how many there are, how each one actually operates, which ones depend on someone rebuilding numbers by hand, and where errors or fraud could slip through.
5. Follow the busiest week. We break down the first days of your close, day by day, to see where the time really goes.
6. Design the future state in four layers: your data brought into one place, one set of definitions, data ready for AI, and answers on demand.

What the Future State Looks Like  

• A governed data foundation. Data comes in raw, gets joined, and is approved before anyone reports on it.
• A reporting layer that rebuilds your critical reports on one definition for each measure, so the numbers match everywhere.
• Controls that move from re-performance to review. Controls that today mean someone rebuilding numbers in Excel become scheduled comparisons. Your reviewer looks at the exceptions. The people who build a number can’t approve their own work.
• An intelligence layer. AI reads your approved data through a read-only, logged connection. It never produces a number of its own.

What You Get  

• A phased roadmap, sequenced by what has to come first, then by the manual work it removes, its effect on your close, and the visibility it gives leadership.
• A SaaS and data-ownership map: what you pay for, what you own, and what you don’t.
• Controls findings, with the automated, preventive controls that would replace manual work.
• An executive presentation to your leadership team.

The full Assessment also includes:

• A board-ready summary.
• A business case built with your own team: a task-by-task estimate of the hours each phase moves from assembling the numbers to reviewing them, so the same team can take on more without adding headcount.
• A phased investment, approved in stages. Each phase is priced at its own approval, based on what the previous phase actually took.
• Running costs and ongoing support, stated up front.
• The principal risks, each with an owner and a date by which we’ll know, and a short list of decisions for your leadership team.

Two Sizes  

Both are a fixed fee, quoted after a free scoping conversation.

How We Build After the Assessment  

• Run it side by side. Tie it out. Then rely on it. Nothing replaces an existing process until it has run beside it for a full close and the numbers match. A manual control stays the control of record until its automated version has done that and the control owner signs off.
• Nothing gets ripped out. The new platform sits beside the systems you already own until it has earned its place.
• Year-end is protected. Build first, rely later.

A Real Example  

For a publicly traded investment firm, the full Assessment ran three months and covered the entire company. We delivered it with our data-engineering partner. My part was the process and controls work: mapping each accounting process with the people who run it, reviewing roughly 120 working files, counting the controls and how each one operates, breaking down the first week of the quarter close day by day, and designing the future controls and the evidence behind them. Together we confirmed the current-state architecture with accounting, IT, and operations, and designed the four-layer future state, with Claude reading approved data through a read-only connection as the intelligence layer. The leadership team received a phased roadmap, a business case built with their own team, and a board-ready summary.

What Happens Next  

Some companies take the roadmap and run with it. Others start with an AI Prototype Sprint on the first opportunity, or move into a phased build. To see what a build can look like, read how a specialty retailer replaced five SaaS subscriptions with one platform it owns.