We spent a decade inside portfolio companies, making them run better with software. Now we make companies AI-native, starting with a world model of how each one really works.
For ten years we ran technology inside private equity portfolio companies. The hard part was never the software. It was that nobody could describe how the company actually worked. So we built Transcript to hear everyone, and the world model to hold what they know.
The understanding lives with the people who do the work, and it only comes out when what they say stays private. Everything we build starts there.
A world model of a company. A snapshot of how work actually gets done is an ontology; a diary of what usually happens is a world model. Ours starts with people, adds systems, learns from outcomes, and as it fills with motion it stops describing the company and starts predicting it.
This is a ten-year programme and research-grade work, including PhD-level research into how a company's world model is built well. Transcript is how it reaches companies of 50 to 500 people, so they decide better and apply AI where it pays.
Budapest, London, Dubai and Singapore. We work with companies across Europe and the Middle East.
Owner-led companies of 50 to 500 people that want to become AI-native now, and private equity firms with portfolios of them. If you need help this quarter, not a thesis, start with a demo.
