think assets.

What we do

Helping you succeed with AI, one expert workflow at a time.

We help a firm's leadership decide where AI is worth investing in, redesign how its experts work with it, show what that is worth to the firm and its clients, and keep it right as AI changes. We begin with one expert workflow: financial due diligence, a relief claim, a valuation. It is advisory work, with no software to buy. We work with the AI you already use.

The evidence The expertframes the task The AIreads it all They meet The AIflags the gaps The expertdecides Better work, by design

How we work

Four steps, and what you end up with.

Each step answers a question, and leaves you holding something. The method behind them has a name, hybrid intelligence: experts and AI on the same decision, each doing what it does best, by design.

  1. Choose where AI matters

    There is more AI on offer than any firm has the attention or the budget to try, and a capability that fell short a year ago may be useful now. We look across your workflows and tasks, and at the capabilities that bear on them, and judge each on three things: the value to clients, how well the task suits AI, and the effort to change. A demonstration is not a business case, so we also take a first measure on a few of your recent files.Where is AI worth it?

    You end up withYour opportunities ranked, and a first change scoped: which workflow, which task, and what to invest in first.

  2. Design the way of working

    A faster draft is not yet better work. Drafts can arrive sooner while senior review takes as long as it did. One draft is accepted too readily, and another, perfectly sound, is rebuilt by hand. So we redesign the work around the AI you already use: who does what, what must be opened and checked, the record that goes on file when someone signs, and the skill each grade needs. We set out the options and what each costs in your experts' time. Your senior people choose, and your people practise the new way on past files.Are the tools we have improving the work?

    You end up withA way of working your senior people chose, written down, with your people practised in it.

  3. Show what it is worth

    Your senior people agree, before anything is run, what sound work looked like on past files. Your experts then work those files with AI, and without. We work the difference through for the quality of signed work, for capacity, and for what you can offer those who rely on your work, with the cost of making the change and of running it counted in. Time saved is capacity, and capacity is not yet value. In a firm, a fixed fee can turn less effort into margin, and hourly billing can turn fewer hours into a smaller bill. In an in-house team, the hours count only once someone decides where they go. So every commercial assumption is written down for your partners or your board to test. Think Assets promises no improvement, no saving and no return. The recommendation can be to adopt, refine, wait or stop.What is it worth to the firm?

    You end up withThe evidence and the arithmetic, in a short paper for your partners or your board.

  4. Keep it right

    The tools keep changing, and so do the habits of the people who use them. A way of working that was right when it was designed can drift without anyone deciding that it should. We stay alongside the firm's leadership: what matters among new capabilities, where to invest next, and whether the way of working still holds. When the tool, the data or the work changes, your own past files are run again to show what moved.What matters next, and does it still hold?

    You end up withA regular review, and your own past files ready to run again, by you or by us.

The split, in each field

Know what to delegate. Know where to think.

Designing the way of working starts with a split. For every workflow there is a right one between the AI and the expert. Three illustrative splits, one from each field.

Financial due diligence on a specialist contractor. Every balance agrees to its schedule. One rests on an estimate.

The AI does

  • Reads every contract and cost report
  • Agrees revenue and contract balances to the schedule
  • Links each adjustment to its document

Together

  • List what rests on an estimate
  • Set each contract's forecast beside how it turned out
  • Show the case for and against each adjustment

The expert decides

  • Whether the estimate was sound when it was made
  • Whether a loss is a one-off or a cost of contracting
  • What the buyer is told before the price is fixed

A research and development relief claim. Every cost agrees to payroll. The engineers' notes say something else.

The AI does

  • Reads the project documents and timesheets
  • Maps costs to categories
  • Drafts the narrative, once the position is decided

Together

  • Test each statement against its source
  • Separate what is asserted from what is evidenced
  • Set each cost beside the work it paid for

The expert decides

  • Whether there was an advance
  • Whether the uncertainty was real
  • Which costs qualify, and what goes on file

A valuation for secured lending. The lease summary was accurate. The break option was in the side letter.

The AI does

  • Reads the leases and side letters
  • Extracts the terms
  • Assembles the comparable evidence

Together

  • Mark what was found and what was not
  • Reconcile the tenancy schedule to the documents
  • Set competing evidence side by side

The expert decides

  • The weight each comparable carries
  • What a break or an incentive does to the value
  • The figure, and the basis it is reported on

See all three worked through, and try each review

What we design and build

Built around your experts.

Where the work needs it, we design and build what helps your experts decide, around the AI you already use. In plain words:

Every statement tied to its source

Every figure in a report, a claim or a valuation linked to the schedule, the note or the letter it came from, and the gaps shown before anything is read.

Your own past files, ready to compare

Similar files from your own record: what was assumed, how it was judged, and why.

A clear view of where the AI can be relied on

Task by task, so your people know where to rely and where to look.

A way to check again when anything changes

Your own past files, run again whenever the tool, the data or the work changes.

What we build for you is yours.

Your people

The skill is part of what we deliver.

A firm's asset is how its people think. A way of working changes nothing until people can work it. So building your people's skill in working with AI is part of every engagement.

When to rely on it

Each person sees, task by task, where the AI earned reliance on your own files, and where their own trust ran ahead of it or behind it.

When to challenge it

People practise the checks on past files, where your senior people have already agreed what sound work looked like, before they use them on live work.

What each grade still does by hand

A skill that is not used fades, and with it the ability to catch the AI when it is wrong. We set out, grade by grade, the work that stays with people, so the judgement behind tomorrow's signatures is still being built.

What changes when AI joins the work, and what can be designed, is set out in The thinking.

Three engagements

Begin with the first. The second is where the work is.

Three engagements cover the four steps. A firm that already knows which workflow it wants to start with can go straight to Design and Evidence.

Step 1 · Where to begin

Adoption Review

Choose where AI matters: which workflows and tasks, which capabilities, and what to invest in first.

You receive

  • Your workflows and tasks judged on client value, how well each suits AI, and the effort to change
  • The capabilities that matter for them now, and those that can be left
  • A first measure on a few recent files
  • Your opportunities, ranked
  • A first change, scoped

Steps 2 and 3 · The core of the work

Design and Evidence

Design the way of working for one workflow, and show what it is worth to the firm and its clients.

You receive

  • The design: who does what, what must be opened and checked, and the sign-off record, chosen by your senior people
  • The people: your experts practised in the new way, and the skill each grade needs
  • The evidence: your own past files worked with AI and without, judged by your own senior people
  • What it is worth: the arithmetic for quality, capacity and what you offer clients, with every assumption stated
  • A short paper for your partners or your board, with a recommendation to adopt, refine, wait or stop

Step 4 · As AI changes

Adoption Partner

Keep it right: independent advice alongside your leadership on what matters, where to invest next, and whether the changes are helping the firm.

You receive

  • A short brief at each review: which new capabilities matter for your work, and which can be left
  • A ranked list of decisions for your leadership: where to expand, redesign, test again or stop
  • A review of how the work is really being done, and whether the changes are helping the firm
  • Your past files run again, by us or by you, when the tool, the data or the work changes
  • A record of what was decided, why, and what would call for the next test

What we need from you

What it takes from your side.

A sponsor, one workflow, a few of your experts, and past files. The evidence comes from those files. What counts as one differs by field, and it is better said now than found out later.

Finance

Past reports, with their data rooms.

The report as signed, and the documents it was built from. Each one is a file to work again.

Tax

Past claims, with their working papers.

The claim as filed, and the client's records behind it. Each one is a file to work again.

Real estate

Past valuations, like for like.

The same kind of property, on the same basis, some since sold or valued again. Each one is a file to work again.

01

A sponsor, and one workflow

Work your experts sign: financial due diligence, a relief claim, a valuation. And a sponsor in the firm's leadership who wants it done well.

02

A few of your experts

Some to work the past files, and one or two senior people to agree what sound work looked like. Their time is agreed in the proposal, before work starts.

03

Time with whoever runs the tools

To see the AI as your people have it. A change to a template or a setting needs an owner on your side, agreed before it is tested. We need nothing from the vendor.

Your files stay in your environment throughout. An Adoption Review needs only a few recent files and a session with the tools.

Who is responsible for what

We are responsible for

Turning your question into work that can answer it. Applying the research. Setting out the options, with what each costs and how it could backfire. A recommendation with its reasons and its limits, including where a change does not help.

Who is responsible for what

You keep

Your senior people say what sound work looks like, and settle disagreements in their own field. Your experts keep responsibility for every professional conclusion. The firm decides what to adopt, and who makes each change to the tools.

For innovation, data and risk teams

You test the tool. We design how experts work with it.

Where a firm has its own AI or data team, we work alongside it. Both tests are needed, and they can disagree. Every field in a schedule can be right and the signed work built on it still wrong.

You test the tool

  • The unit is a field, a prediction or a run
  • It is measured in fields right, drift over time and cost per run
  • It tells you how the AI performs

We work on the expert's decision, with your experts

  • The unit is the report, the claim or the valuation an expert signs
  • It is measured in where the expert followed the AI, where they overrode it, and what the difference is worth
  • It tells you how your experts perform with the AI, and without
  • You keep the past files, the scoring rules and the tools
  • You can run them again on every new version
  • Every step is written down for your team to check

Questions before you begin

What people ask us first.

We already have an AI or data team. Where do you fit?

Alongside it. Your team tests the tool: whether a field, a prediction or a run is right. We work with your experts on the work a person signs, and on how the two fit together. Everything we build stays with you.

Do we need to buy a new AI tool?

No. We start with the AI you already use, and we resell no one's product. If the evidence says a different capability would help, we say so, and the choice is yours.

How much of our people's time does it take?

It is agreed in the proposal, before work starts: who takes part, in which sessions, and what your senior people are asked to decide. A count of heads is not an estimate of time, so we set out both.

Does everyone in the team need to take part?

No. A few of your experts work the past files, and one or two senior people agree what sound work looked like. The way of working that comes out of it is for the whole team.

What if the new way of working does not help?

Then the paper says so. The recommendation can be to adopt, refine, wait or stop, and each comes with its reasons.

Who is responsible for the professional conclusions?

Your experts, as now. We design the way of working and the test. Your people sign the work, and your firm decides what to adopt.

What do you see of our files?

Your files stay in your environment throughout, and access is agreed at the start. We work only with files your own terms allow you to use. The first conversation needs no documents at all.

Can our own team run the test again?

Yes. The past files, the answers your senior people agreed and the scoring rules stay with you, written down so your team can run them on every new version.

How it starts

Three steps to a first engagement.

  1. The pack for your field

    Ten pages, for finance, tax or real estate, or tell us your field. Ask for the pack for your field.

  2. Thirty minutes with our founder

    Bring one workflow, with no client details. We sketch where AI could matter in it, and how that would be shown.

  3. A proposal

    The scope, the fee and what the work can show, all agreed before you commit.

Tied to no vendor. We resell no one's product. Fixed fees, agreed before we start.

Start here

Get the pack for your field.

Ten pages. Pages 2 to 4 take five minutes.

Get the pack for your field

Thirty minutes with our founder, not a salesperson. Bring one workflow, with no client details, and we will sketch where AI could matter in it.

Arrange the 30 minutes