think assets.

AI advice for professionals

Succeed with AI, by design.

Think Assets helps professional firms and in-house teams decide where AI is worth investing in, redesign how their experts work with it, and show what it is worth. Research-led advice, tied to no vendor.

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

Where things stand

Pressure to move. A need to know what works.

Every professional team has capable AI now. Moving too fast builds in weak assumptions. Waiting leaves useful capability unused. Three questions are left open.

Where is AI worth it?

Demonstrations are plentiful. Senior attention and budgets are not. Which workflows, which tasks and which capabilities deserve the next investment?

We help you choose

Are the tools we have improving the work?

Drafts arrive faster, yet senior review can take as long. In some places a draft is accepted too readily. In others, sound work is rebuilt by hand.

We redesign the work

What is it worth?

Time saved is capacity. It has value only once someone decides what to do with it. What to offer those who rely on your work, how to staff and where to invest next should not rest on impressions.

We show what it is worth

How we help

One expert workflow at a time. Chosen, designed, shown.

We start where the value is: the work an expert signs. Four steps take one workflow from an open question to a way of working the firm can stand behind.

  1. Choose where AI matters

    Which workflows and tasks, which capabilities, and what to invest in first. Judged on client value, how well the task suits AI, and the effort to change.

    You end up withYour opportunities ranked, and a first change scoped.

  2. Design the way of working

    Who does what, what must be opened and checked, what goes on file when someone signs, and the skill each grade needs.

    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 experts work your own past files with AI, and without. The difference is worked through for quality, for capacity and for what you can offer those who rely on your work.

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

  4. Keep it right

    The tools and your people's habits keep changing. We stay alongside your leadership: what matters among new capabilities, and whether the way of working still holds.

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

How each step works

What success looks like

Extend what AI enables. Strengthen what your experts do.

AI used in the right places, in the right way, with trust that matches the evidence. Too much trust is a risk. Too little is a waste.

A stronger service

AI widens the evidence your people can consider and the alternatives they can explore. Those who rely on your work get deeper advice, more clearly explained.

Experts on the work that needs them

Your people delegate what the AI does well, and put their judgement where it does not.

Capacity you can use

Preparation, checking, rework and senior review, counted together. Time released has value when your team can use it.

People skilled with AI

Each grade knows when to rely on the AI, when to challenge it, and what it still does by hand.

We promise no improvement, no saving and no return. We set out the evidence and the arithmetic on your own work, and you judge it.

Where it is won or lost

Faster is not the same as better.

Three pieces of signed work, one from each field. In each, the AI's draft agreed with its source, and the judgement turned on something else. A firm succeeds with AI when its way of working reaches that point every time.

Try each review yourself

Follow an in-house team through one engagement

Due diligence report · a specialist contractor

Adjusted EBITDA

As drafted with AI£3.00m
Once the source is opened£2.64m

A review after the year end raised the cost to complete the largest contract by £450k.

£2.16m of enterprise value, at six times

R&D relief claim · a new pricing engine

Costs in the claim

As drafted with AI£410k
Once the source is opened£246k

The engineers' notes show about 40% of the cost was data migration. £246k is a ceiling: whether the rest qualifies is a judgement.

£164k of the claim is data migration

Valuation · a multi-let office

Weighted lease term

As drafted with AI7.4 years
Once the source is opened5.2 years

A side letter in the data room gives the largest tenant a break at year three of its lease.

£440k a year of rent can leave

Illustrative. Every figure in each draft matched its source. Work through each one

Why Think Assets

Three kinds of knowledge, rarely found together.

The work itself

Expert work as it is done, with a background in finance, tax and real estate: senior review, the training of juniors, and what a signature commits a firm to.

For the design

How experts and AI work well together

Grounded in peer-reviewed research on testing AI systems, calibrated confidence and how people and AI decide together.

For the evidence, and keeping it right

The economics of expert work

What those who rely on the work value, what it costs to deliver, and what a change means for capacity and for the people who do the work.

For what it is worth

Built around your experts. Where the work needs it, we design and build what helps them decide, around the AI you already use and inside your own environment: every statement tied to its source, your own past files to compare with, and a way to re-test when anything changes. We build your people's skill with it too: when to rely on the AI, when to challenge it, and what each grade still does by hand. What we build for you is yours.

Who is behind Think Assets

  • Tied to no vendor
  • We resell no one's product
  • Your work stays in your environment
  • Fixed fees, agreed before we start

Insights

Think through the real work.

Designing the work

Design the work, not just the prompt

How evidence, AI and professional judgement are put together decides what a firm gets from AI.

Read the article

Confidence

Confidence should follow the evidence

Too much trust in AI is a risk and too little is a waste, so test what people believe and what they rely on separately.

Read the article

Keeping it right

An AI trial is a snapshot, not a permanent verdict

When to look again at what is delegated and what experts review, as the tools, the sources and people's habits change.

Read the article

Start here

Get the pack for your field.

Ten pages. Pages 2 to 4 take five minutes.

  • The decision and the grey area
  • Why the obvious fixes fail
  • The way of working we would design, with a sample sign-off record
  • What it means for value and for your people
  • How we would show it on your own work, and the sources

We use your email to send the pack and one short follow-up. Reply "stop" and you will hear nothing more. Privacy notice.

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