The data layer
for opaque markets.
We build and run AI-driven platforms for property, emerging markets and capital networks. They are all places where the information you need is spread across too many sources, out of date, or simply not public.
The problem
Most markets are not inefficient. They are just undocumented.
Think of a property market with no comparable pricing. A region where nobody publishes reliable supply data. A funding round that only happened because two people already knew each other.
None of these are short of money. They are short of a record. So we build the record first, and then the tools that read it.
Where we work
Three markets, one underlying problem.
Property platforms
Listings, pricing and provider data pulled together from hundreds of sources that rarely agree with each other. You can search, compare, and see the market context all in one place.
Emerging market intelligence
Coverage of the regions where good information is hardest to come by. We track supply, pricing and sector signals, then use our own models to flag what changed and what it might mean.
Deal flow and capital intelligence
Deal tracking, portfolio oversight and founder-to-investor matching for angels, VCs and networks. Built around the relationships that private markets actually run on.
How it works
From unusable sources to a decision.
Collect
Public sources, partner feeds and market-specific data. Most of it inconsistent, incomplete, or impossible to compare side by side.
Structure
We normalise all of it into clear models and categories, so records become searchable, filterable and genuinely comparable.
Analyse
Applied AI and research methods across the structured layer, looking for the patterns, gaps, risks and movement worth knowing about.
Connect
The result becomes something you can act on. Find the property, read the region, track the deal, reach the person.
How we build
Self-funded, so the work is the only thing to answer to.
We keep what we build
We build, run and maintain our own platforms. The people writing the code are the people carrying the pager, which tends to focus the mind.
We specialise
Complex data, and the systems that make it useful. We are not going to pivot toward whatever happens to be fundable this year.
We show our working
Assumptions stay where you can see them. A model that cannot explain its own drivers is not finished yet.
AI where it earns its place
We reach for it to improve search, matching, modelling and analysis. We leave it well alone where it cannot safely decide.
- Own what you ship
- Specialise
- Show your working
- Explainable models
- Provenance on every figure
- Decision support, not verdicts
- Built to operate for years
Published research
Research that ships.
Our whitepapers are not marketing. They cover the problems we are working on, the assumptions we are testing, and where we think the technology is heading. We write them for people who have to build or fund something similar.
Working on something similar?
We are always up for a conversation about data, distribution, or any of the problems above. We are not an agency though, and we do not take client work.