High-stakes engineering, operated tightly
We are the engineers other engineers call when it cannot fail.
A closed group of principal engineers and CTOs for rescue, 0→1 builds, and app reviews when the board, deadline, or incident room needs evidence instead of optimism.
Operating loop
- 01 Frame the risk: system map, owner map, decision log.
- 02 Stabilise the path: ship the smallest reversible move.
- 03 Leave evidence: review packet, runbook, next-call memo.
Artifacts first. Meetings second.
01 — What we take on
01 — What we take on
Rescue & rearchitect
Your system is on fire. We've put out worse.
Legacy that nobody dares touch, a migration that stalled, a platform buckling under its own success. We go in, stabilise, and rearchitect without stopping the business.
Evidence object
- Failure map: what is broken, what is merely noisy, what must not move.
- Stabilisation branch with rollback plan and production runbook.
Build 0 → 1
From whiteboard to production, with senior hands only.
New product, hard deadline, no room for a learning curve. We design the architecture and ship the first versions ourselves — no juniors practising on your runway.
Evidence object
- Architecture brief tied to first-user workflows, not abstract preference.
- Production slice with observability, deploy path, and ownership notes.
App review
You shipped fast. Now you need the risks in plain sight.
A fixed-scope review for AI-built or fast-moving apps: architecture, security, reliability, product gaps, and the next technical decisions that actually matter.
Evidence object
- Risk register across architecture, security, UX, data, and delivery.
- Prioritised remediation plan that separates taste from risk.
02 — Selected work
Hard problems. Shipped.
Client names withheld. We'll walk you through the architecture and references under NDA.
03 — The seats on your build
No account managers. The people who build it.
Jonathan Koch
Technical Architect
STRATEGY · ARCHITECTURE · EXECUTION
A war-time CTO who still ships code — infrastructure to interface, strategy to execution. The seat you want when the call is hard and the standard can't move.
Daniel Amireh
Full-Stack Engineer · Ruby/Rails
RUBY · RAILS · DEVELOPER TOOLING
Full-stack Rails and the developer tooling around it — the plumbing that makes a team ship faster. Berlin-based, and the work is public.
Christian Rühringer
Systems & Web Engineer
TYPESCRIPT · C · SYSTEMS
From Linux drivers in C to a Quake engine in TypeScript — at home from the metal to the browser. The work is public; check the repos.
Principal Engineer · Security
CRYPTO · REGULATED DATA · AUDIT
Designs the trust boundary and the audit trail so the system clears review the first time, not the third. Built for the regulated case from day one.
Staff Engineer · Platform & Delivery
KUBERNETES · CI/CD · ZERO-DOWNTIME
Owns the path from commit to production: migrations that don't take the system down, releases you can reverse. The boring discipline that keeps the lights on.
Named members lead — the rest we share under NDA. Ask in the intake for the full CVs.
04 — How we operate
You get a working loop, not a slide deck about one.
Every engagement is run around visible evidence: the current model, the decision under pressure, the smallest safe move, and the artifact the next team can inherit.
Typical docket
- Architecture map with known unknowns and owners.
- Risk ledger sorted by impact, reversibility, and blast radius.
- Shipped remediation path with runbook and next decision memo.
We don't publish testimonials we can't attribute. Get to a fit and we'll put you on a call with people we've shipped for — direct, under NDA.
05 — Request an engagement
We're selective — so this form is too.
This goes to a partner, not a sales inbox. If it's a fit, you'll hear back within 48 hours. If it isn't, we'll tell you straight.