Infrastructure
Data, compute, and the systems engineers build on. Long sales cycles, deep moats, and a founder profile we recognize.
At pre-seed there is rarely enough evidence to be right for the correct reasons. So we bet on judgment, speed, and the specific unfair advantage a founder has already earned.
Pre-seed markets are guesses. A fund that only invests in one sector eventually funds a mediocre team inside its thesis and passes on a great one outside it. We would rather stay generalist and be honest about what we do not know.
The companies that break out decide how they will reach people in the first month, not the first year. We ask about it early because retrofitting distribution is expensive.
We raise deliberately small funds. That means one company returning the fund is enough, so we never have to push a founder toward a round they do not need.
A fundraise costs a founder attention they cannot get back. We move in two weeks so the cost stays contained either way.
Generalist mandate, four areas of real depth.
Outside them we still invest, and we say plainly that our help will be about company building rather than the domain.
Data, compute, and the systems engineers build on. Long sales cycles, deep moats, and a founder profile we recognize.
Models doing real work inside a workflow. We look for domain knowledge first and model choice second.
Payments, treasury, and the operating layer under small businesses. Unglamorous and durable.
Identity, runtime defense, and the surface area that agents are creating faster than anyone can patch it.
You know something about this problem that most people cannot learn from reading. Usually it comes from having lived inside it.
How much better the team got between the first meeting and the second. This predicts more than any metric available at pre-seed.
Even one. Someone who would notice if the product disappeared tomorrow is worth more than a waitlist of thousands.
Founders who can say what they do not know without hedging tend to make faster decisions later.
Not every good product needs venture money. We ask what breaks if it stays small, and we respect the answer either way.
Saying this in public saves everyone a meeting.
If the round is closing next week and you need a name to fill it, we will be too slow and too curious.
A market thesis with the engineering outsourced does not survive the first year.
We admire it. We are not the right first check for a company that needs a factory before it needs a customer.
If the business only works when the customer does not understand it, we pass.
Tell us what you are building and why you are the person building it. Two paragraphs is plenty.