For the people whose name goes on the work

AI now drafts the work. Your experience decides whether it is right.

AI Elite Thinkers helps experienced professionals make AI a working part of the job they do well. The tools are open to anyone. The edge goes to those who can tell a fluent draft from a right one.

Start freeTry the check on a sample draft

AI now drafts a board summary, a contract review, or a page of variance commentary in minutes.

The drafts read well. Some of the figures in them trace to nothing.

Checking one still takes as long as it did before the tools arrived.

The name on the work is still yours, and so is the explanation when a figure turns out to be wrong.

So the weight of the job now sits with whoever reviews the draft.

You have spent years deciding which figures, clauses, and claims to trust, and every AI draft now needs someone who can.

These professionals are not behind. They have what the tools lack: judgement about which answers to trust.

The AI Playbook, p. 4

Check the draft yourself.

This is the kind of summary an AI tool writes from a board pack, ready to paste into the papers. Beside each figure is where it came from, or a flag where nothing in the pack supports it.

Sample: an AI draft of a board summary, figures invented for the demonstration

Group revenue for the quarter was £14.2m (source: Board pack p. 4, cell B4), 6.1% (source: Board pack p. 4, cell D4) ahead of budget, led by the two new contracts in the Northern region. Gross margin held at 38.4% (source: Management accounts p. 2, cell F18), comfortably above the sector median of 34% (source: UNSUPPORTED: no source in the documents you supplied) for mid-market services businesses. Operating cash flow of £2.3m (source: Cash flow statement p. 6, cell C12) kept the business well inside its banking covenants. On the current pipeline, revenue should grow at 12% (source: UNSUPPORTED: no source in the documents you supplied) a year through 2028, which supports the case for the second site.

  1. £14.2mTraced
  2. 6.1%Traced
  3. 38.4%Traced
  4. 34%No source
  5. £2.3mTraced
  6. 12%No source

4 of 6 figures traced. Two have no source.

The two without a source are a sector benchmark and a growth forecast: the figures a board is most likely to repeat.

The real check is a prompt from The AI Playbook (p. 45). Run it in the same conversation where you gave the model your documents.

Show me where each figure came from

Before I use this draft, list every number, name, and date in it.
For each one, give:
- its source: the document and page or section I gave you, or none, and
- whether the source states it, or you calculated or inferred it.
Mark anything you cannot point to a source for as UNSUPPORTED.
Do not correct or soften anything yet; I will decide what stays.

Treat the list it gives you as a map of where to look. Open the passages behind the figures that carry your conclusion, then decide what you sign.

Build AI into the expertise you already have.

AI Elite Thinkers is a way of working: AI inside the real job, held to the standard your name already carries.

  • Start from the real work

    Month-end close, board packs, contract review, long documents, and competitor research, each worked through with the prompt written out in full.

  • A check for every draft

    Each workflow comes with what to verify before the work leaves your desk, so the judgement stays with the person who signs.

  • Built by a practitioner

    Jamie Whitson, a PwC-trained auditor and former Head of Finance, founded AI Elite Thinkers and re-checked the Q4 2026 editions against the vendors’ own pages in October 2026.

Start where you are: from first setup to Claude run as a system.

Begin free with a map of the tools, a proper Claude setup, or 5 finance workflows. Go platform-deep when you are ready, paid once and yours to keep.

Free

The AI Playbook

A map of ChatGPT, Claude, and Gemini for experienced professionals, in 51 pages with 14 prompts to copy.

Free

The Claude Quick Start

Set Claude up properly in 45 minutes and give it your context, with no coding, in 50 pages.

$47

The Claude Playbook

Long documents, contract review, variance commentary, and board packs, in 202 pages with 90+ prompts.

$47

The ChatGPT Playbook

GPT-6, deep research with citations, and data analysis you can check, in 271 pages with 117 prompts.

$47

The Google AI Playbook

Gemini across Gmail, Docs, Sheets, and Drive, with Gemini Notebook for research, in 195 pages.

$97

The Setup

A folder your own Claude reads first. It drafts your monthly commentary from your own numbers and ties each figure to a cell.

Paid once, 30-day refund.

Each paid guide has a 30-day refund; the bundle has 60 days, with conditions.

Jamie Whitson, PwC-trained auditor, former Head of Finance

Jamie Whitson

PwC-trained auditor, former Head of Finance

I served my articles at PwC, auditing property, logistics, and retail businesses, then spent 9 years as Head of Finance at a family-owned entertainment group.

An auditor traces each number to a source and raises anything that does not trace. I started AI Elite Thinkers to bring that habit to the new tools, so experience counts for more, not less.

Founder of AI Elite Thinkers

More about Jamie

Bring AI into one piece of real work this week.

The AI Playbook if you are weighing up the tools, The Claude Quick Start if Claude is new to you or gives you generic answers, or The Claude Finance Runbook if you want to hand it real finance work. Each one opens as soon as you enter your email, and a copy goes to your inbox.

Start free