Atlas
Fintech / data-opsAI-assisted data-ops layer for small finance teams. First paying customers; revenue still in pilot range.
MVP live
Not-a. is what comes after the venture studio. Closed format —
we originate, build, and operate every company ourselves. No
spin-outs11 / spin-outswhen a studio hands a project
to outside founders and takes
equity. we don’t. the same team
operates every company we start.,
no outside founders. Whiteboard to revenue in our own hands.
Unit economics decide whether a project lives or closes.
AI-native execution22 / AI-nativethe reason a team this small
ships at this pace. AI compresses
what used to take ten people across
product, engineering, design, and ops
into work the founding team does directly.
is the engine — speed and leverage without scaling headcount.
We don’t chase
unicorns33 / unicornscompanies valued at $1B+.
unicorn-chasing requires burning capital
for years on the chance of one
outsized exit. we optimize for the
opposite — companies that pay back
quickly and keep paying.;
we build predictable companies that return their multiple, with the
option to flatten growth and run profitably any day of the week — an
off-switch44 / off-switchstop growth investment at any point;
the company runs profitably as-is.
most venture-backed companies can’t
do this — their economics only
work at scale. ours work at any size..
Capital joins where the math already works. Both sides see the same numbers before the conversation starts.
The old studio format was designed for a different decade — when
capital was cheap, teams were large, and
Series B economics11 / Series Bthe valuation logic where a
company needs 10×–100×
growth potential to justify years
of burn before profitability.
default operating assumption of
the last decade. no longer reliable.
justified anything. None of those conditions hold anymore.
What used to take twenty people now takes four. What used to require years of burn to look interesting now needs to pay for itself by month six. The studio that doesn’t restructure around this disappears.
Not-a. is restructured. Fewer people, faster cycles,
public economics22 / public economicswe publish revenue, costs,
and unit-economics for every
Not-a. project — internally to
the team, externally to anyone
watching. transparency operates
as a discipline, not a marketing layer.,
and the discipline to close what doesn’t earn. Our position is
published — The Not-a. Manifesto lays out what we built.
18 Sins of Venture Studio explains what the old model got wrong.
The shape of our work is recognizable. We source ideas, run marketing tests before code, build MVPs, ship, measure, iterate. We have a playbook and we follow it. The discipline is conventional.
AI is not a tool we use. It is the reason a team this small can ship at this volume. Research, build, design, ops — all run through it from day one. The cost structure that follows is not an optimization of the old model. It is a different model.
Four operators. We don’t grow into ten. We don’t migrate into the projects we start — the founding team operates every company, in parallel, without spinning out into them. This is the structural decision the rest of the model depends on.
We do not chase unicorns. We are not against them — they are simply not what we optimize for. Each project is built to reach unit-economics breakeven before anything else. Some compound. Some get sold. Some get closed. None get carried by faith in a future round.
Most studios start with a product and look for an audience. We start with the audience. Every project begins with building a community around the topic — before the MVP, before the landing page, sometimes before the idea is final. Pivots become cheap. Distribution is owned, not rented. Performance channels are the last resort, not the first.
What is being built, what is working, what is being killed, where capital goes — published. To the team always; to investors as terms; to the public as much as we can. Reporting is not a quarterly formality. It is the discipline that keeps the rest honest.
A snapshot of what is moving through the studio. We keep three to five in parallel — enough variety to spread risk, small enough that nothing rots in the queue. Stages move; this list updates when reality changes, not on a schedule.
AI-assisted data-ops layer for small finance teams. First paying customers; revenue still in pilot range.
MVP liveShort-form, skill-based micro-coaching app. Strong organic loop; monetization currently in test.
Early tractionVoice-first reading companion for children. Closed beta; unit economics not yet readable.
ResearchAsync family-logistics tool. Peeled off Q1 2026 with its own team after clear PMF signals; now a standalone company.
Spun out · 2026Financial tooling for gig-workers. Graduated in 2025, now operating independently with its own investor base.
Spun out · 2025B2B scheduling infrastructure for service teams. Acquired by a US scheduling platform; proceeds returned to the LPs.
Acquired · 2023Lightweight observability for small engineering teams. Acquihired into a larger DX platform the following year.
Acquired · 2022AI calendar assistant. Retention flat, unit economics never closed — we pulled the plug before burning more.
Shut down · 2025Community app around shared hobbies. Organic loop stalled after the first cohort; the off-switch did its job.
Shut down · 2024Most studios negotiate every deal from scratch. We don’t. Our share split, fee structure, and exit mechanics are written down — the same numbers for everyone, every round.
If you allocate to venture studios, the whitepaper is the entry point. Read it first; if the model fits your thesis, the rest is paperwork.
Long-form essays and short notes from the studio — on what we are building, what we are reading, what we got wrong. Cadence is irregular on purpose: we write when there is something to say, not every Tuesday.