News How we work

Four signed engagements in six weeks, and the gap we started in.

The enterprise AI services firms start at enterprise scale. Most of the property market sits underneath it.

Teddy James · · 4 min read

Tercero Analytics is live. Four signed engagements in the first six weeks: a UK industrial owner, a US multifamily sponsor, a European private equity real estate firm, and a UK industrial agency. Four different mandates and one brief. Institutional grade AI work for funds and advisors who should not have to buy an enterprise contract to get it.

The week the fourth was signed, Anthropic announced a $1.5bn raise from Blackstone, Goldman Sachs and Hellman & Friedman to launch an enterprise AI services firm. The model is forward deployed engineers embedded inside Blackstone’s portfolio companies. Minimum viable client: enterprise scale.

That is useful confirmation that the category is real. It is less useful if you are the boutique investment manager in Mayfair with £300m under management. Or the family office running a £150m mixed portfolio. Or the regional fund manager who needs counterparty intelligence on a £40m industrial deal by Thursday.

That gap is where we started.

What the work actually is

Variance to budget analysis. AGM and LP reporting. Peer benchmarking. Counterparty intelligence. Covenant workflows. Built bespoke, deployed in weeks, sized for the boutique end of the market.

None of that is a product you log into. It is the work a firm already does, done with a system built around the way that particular firm holds its data. A £300m manager and a £2bn manager do not have the same problem at different scales. They have different problems, and the smaller one is usually the harder engineering job, because there is less structure to work with and nobody in-house to maintain what you leave behind.

“The question is not whether AI changes how you operate. It is who has actually built something for your scale.”

Teddy James, Tercero Analytics

The data layer underneath it

Every engagement sits on the same corpus: more than 35,000 commercial property transactions worth over £3 trillion, 4,400 fund profiles, 3,400 debt facilities, and 42,000 REIT earnings call question and answer exchanges across the UK, Europe and the US, all embedded and searchable.

That last one is the piece most people underrate. It is the record of how a CFO has answered the same awkward question across twelve consecutive quarters, and what changed in the wording the quarter before the guidance moved. That is available to anyone who wants to read four years of transcripts. Almost nobody does, because it takes a week and the call is on Tuesday.

Who this is for

If your firm sits between £100m and £2bn under management, you are in the band this was built for. You have the same analytical problems as a much larger house, a fraction of the headcount to throw at them, and no appetite for a two year platform programme that ends in a login and a training deck.

What you get instead is a short data review, a scope written against what your data actually looks like, and something running in weeks. If that sounds narrow, it is deliberately narrow. It is the only version of this work we have found that survives contact with a real portfolio.