Applied AI
Domain-specific systems where output can be graded against ground truth. Underwriting, clinical documentation, industrial inspection, legacy code translation.
We avoid: thin wrappers whose only moat is prompt engineering.
Everything below is written down so you can decide whether to spend an hour with us before you spend it.
Market size is a lagging indicator. We start from a problem that already costs someone real money every single day, then ask whether technology has only recently made it solvable.
Domain-specific systems where output can be graded against ground truth. Underwriting, clinical documentation, industrial inspection, legacy code translation.
We avoid: thin wrappers whose only moat is prompt engineering.
Ledgers, reconciliation engines, embedded payments, treasury tooling, and compliance systems sold into regulated institutions.
We avoid: consumer neobanks competing on incentives.
Grid software, building electrification, freight optimisation, and process heat. Businesses whose unit economics work without subsidy by year five.
We avoid: carbon offset marketplaces.
Often before incorporation is finished. We can be the only institutional name on the round, or write alongside angels.
We lead roughly two-thirds of our seed rounds, take a board seat only when asked, and always take an observer seat.
Reserved capital deployed through Series B. We support our companies in flat and down rounds, not only in good news.
We designed this backwards from the founder's calendar. Every stage has a deadline we hold ourselves to, and you always know which step you are on.
We do not run "tracking" meetings for companies we have no intention of funding.
Start the processA form or an email. Two paragraphs is enough; a deck is optional. Response within five business days.
Forty-five minutes, camera optional, no formal presentation required. We will have read whatever you sent.
A working session with your team, a technical review, and two customer references you choose.
Full partnership vote every Monday. You hear the outcome, with reasoning, the same day.
Every fund claims to add value. Here is the specific, finite list of things we do.
A full-time talent partner who sources your first ten engineering and go-to-market hires, not just a shared job board.
Warm introductions into our network of 400+ operators at banks, insurers, utilities and manufacturers.
We build your Series A narrative six months early and introduce you to the ten funds most likely to say yes.
Co-founder disputes, layoffs, failed launches, down rounds. We have been through all four and will pick up the phone.
No. Roughly a third of our Fund IV investments came through the form on our contact page from founders we had never met. Every submission is read by a partner, not filtered by an analyst.
At pre-seed, none. We have funded companies with a prototype and two design partners. At seed we usually want either early revenue or evidence that a small number of users would be genuinely upset if you shut down.
Occasionally, but be aware of the trade. Outside our focus areas our money spends the same, and our help is worth considerably less. We will say so plainly rather than overstate what we can offer.
Never at pre-seed. At seed, only when we lead and the founders want one. We always take an observer seat so we stay informed without adding governance overhead to a six-person company.
Europe, North America, and Southeast Asia, from offices in Lisbon, Austin and Singapore. We are comfortable with fully remote teams provided there is a legal entity we can invest into.
If you have got this far, you already know whether we are the right fund for you. Take five minutes and tell us what you are building.