First note. Hello from the lab.

This is the first entry. It is going to be short, but it will hold up.

Xipher Labs is two engineers and a network of senior operators who get pulled in by name. We pick a few problems a year and build systems around them — product, infrastructure, Web3, AI when it earns its place, and the kind of security review most teams need but rarely have full-time.

We did not start the lab to scale. We started it because the projects we wanted to ship needed a home with the right defaults: senior people, real specs, code review that means something, and security as a first-class concern. If you have worked in a place where those defaults were missing, you know the cost.

What this page is for

A small public set of notes. Not a content strategy. Not a newsletter funnel. The pace will be slow on purpose — closer to Patrick Collison's writing rhythm than to the LinkedIn cadence everyone else has adopted.

Three kinds of writing live here:

• Field notes from active projects, when we can say something useful without breaking trust.
• Short essays on how we think about engineering, AI, sovereign data, and security.
• Retrospectives when something we shipped breaks, or when we change our mind. Especially when we change our mind.

We are not going to publish unless we have something worth your time. The page will be quiet for stretches. That is by design — Jason Fried and David Heinemeier Hansson made the same argument for years in It Doesn't Have to Be Crazy at Work, and we agree.

What we are building right now

The current state of the work lives on the home page under "What we build". Short version:

• Licit.ar — an AI-powered tender aggregator. Pulls public procurement notices from every Argentine government source into one searchable feed.
• BioVault — sovereign medical data infrastructure. Patients control everything; the record is accessible when it is most needed, with no privacy compromise.
• Membership OS — a permissionless membership system for Web3 communities. Proof of work across categories, reward distribution by contribution.
• Walter OS — our internal agentic operating system. The discipline we apply to every project.
• Confidential security work — reviews and audits for clients we can't name. The work that doesn't ship with a logo.

If any of those sound like a problem you want to talk about, the contact form is the right place.

What we believe

Two things, mostly.

Most of what gets built doesn't need to exist. We are picky about which problems we take on. A clear stakeholder, a real constraint, an outcome that matters when shipped. Everything else can wait or be done by someone else. This is closer to Marty Cagan's framing in Inspired than to the "ship fast, iterate" school — both work, we picked our side.

The boring parts are load-bearing. Specs, plans, code review, observability, runbooks. Teams that take those seriously ship faster in the medium run, even when it feels slower at the start. We are betting our calendar on that.

What you can expect from us

Short notes. Real opinions. No hype. We will tell you when we are wrong.

If you want to know when something new lands, the quiet newsletter sends one short message per essay. No drip, no funnel.

Welcome to the lab.

Further reading

• Working in Public — Nadia Asparouhova on how small teams maintain disproportionate output.
• Shape Up — Ryan Singer / Basecamp on the 6-week cycles that influence how we scope projects.
• The Mom Test — Rob Fitzpatrick on talking to operators before writing code.

>  end of article · v0.1 · 2026