About — A digital foundry 05 / 06

Before the
valuation, the
foundation.

re:MAK3 was started in 2023 by four engineers who had spent a decade building infrastructure for other people's ideas — and grew tired of watching capital follow the loudest pitch while infrastructure followed capital. We thought the order should be reversed.

FoundedFebruary 2023
HeadquartersDistributed — Portland & Lagos
Members24 engineers, 4 partners
StatusPrivately held, employee-owned
I — How it started

The chaotic race for
valuation often skips
the most critical step.

Silicon Valley rewards the promise of disruption over the stability of execution. We'd built four startups between us and watched the same pattern play out each time: capital arrives, the team is told to scale, the infrastructure is improvised — and a year later the engineering org spends every cycle paying down the resulting debt.

The thesis was simple. Most early companies don't need millions in pre-seed cash; they need organised development, scalable compute, and tested infrastructure. The barrier isn't money, it's access.

So we built a foundry to provide that access. Consultancy revenue funds the compute, the compute funds the Ventures, and the Ventures fund the Initiative — and the Initiative keeps us honest about why we built any of it. The loop has held for three years.

II — Five tenets

What we hold,
even when it costs us.

01

No deck culture

We write prose — memos in, memos out. Decks signal motion, not thought, and we've found that the discipline of paragraphs filters out the work that would have failed anyway.

02

Compute as currency

Cash is fungible and forgettable, but a reserved GPU is a real, finite, allocable thing. The whole foundry runs on that distinction, and it forces honesty about what we can actually take on.

03

Embedded, not advisory

We don't sit on boards, hold office hours, or hand over playbooks. We work inside your repo, commit code, leave a runbook — and then we leave.

04

Renewable by design

When a project ends, hardware is wiped and reallocated, and the engineers carry the lessons forward. Nothing is sunk — and the Initiative funds itself from the metabolism of the rest of the work.

05

Public-good, not public-relations

The Initiative work isn't a brand asset, so we don't write press releases for it. The line between paying clients and pro-bono is invisible from the inside — same engineers, same standard.

III — The people

Twenty-four
engineers,
four partners.

We hire slowly. Every member has shipped a system that mattered, and every member commits time to the Initiative each year. Compensation is partner-equivalent across the firm, and the only hierarchy is the one the work demands.

Mara Holst

Mara Holst

Founding partner · Compute
Portland
Idris Okafor

Idris Okafor

Founding partner · Engineering
Lagos / NYC
Yuki Tanaka

Yuki Tanaka

Partner · Ventures
Tokyo
Sasha Reyes

Sasha Reyes

Partner · Initiative
Mexico City
Tomas Brandt

Tomas Brandt

Principal engineer
Berlin
Lina Park

Lina Park

Principal engineer
Seoul
The order of operations
“In order to demand the future we deserve, we must first build it — everything else, the financing, the press, the optionality, is downstream of that.”
Operating doctrine, §1
IV — Continue

Three pillars.
Pick the
one that fits.

Each pillar is a different door into the same foundry — read the one that matches your problem, or open a channel directly.

For enterprisesConsultancy — embedded engineering, audits, founding-team augmentation
For foundersVentures — compute and engineers in exchange for equity
For the public goodInitiative — pro-bono work for schools, NGOs, civic infrastructure
Open a channel → Back to home