About

We sell the determinism your cluster is missing

Cloud vendors sell you somewhere to put things. Platform vendors sell you software and leave. We sell the part neither of them will: a named engineer who builds the cluster, proves it survives failure, teaches your team, and then answers the phone at 3am.

Why this exists

Ask anyone who has run infrastructure for a while where the pain is, and it is never the licence. It is the upgrade nobody dares perform, the etcd backup nobody has restored, the failover that was never actually triggered, the CRUSH map that rebalances during business hours, the GPUs sitting allocated but idle.

Those problems are not solved by a platform. They are solved by people who have seen them before and who are accountable for the outcome. That is the entire premise of this company.

What we are not

  • Not a reseller. We take no margin on hardware or licences. Our incentive is to give you the smallest stack that solves the problem.
  • Not a cloud. Your data stays on your infrastructure. We do not proxy traffic and we do not ship your telemetry anywhere it does not need to go.
  • Not a body shop. We will not sell you a warm body to fill a seat on a spreadsheet. If the right answer is that you need to hire, we will say that.

Where we work

Based in the Bay Area, working with customers across North America and Europe, remotely and on-site when the work genuinely requires hands on hardware. Our own systems run Debian, nginx and systemd, with atomic releases and automated rollback — because we would rather be held to the same standard we sell.

Position statements

Three things we will not do

Every firm has an ethics page. This one is specific enough to cost us money, which is the only kind worth reading.

01

We will not trade vendor lock-in for a bigger invoice.

Everything we deliver is built on open standards — CNCF, Linux Foundation, Apache ecosystems. If you want to take it in-house or move it elsewhere, nothing we build stops you. That is a constraint we accept on purpose.

02

We will not promise an SLA we cannot hold.

If we can do 99.9% we write 99.9%. If something needs a third party, we mark it "requires partner" instead of quietly including it in the definition of uptime. A number you cannot trust is worse than no number.

03

We will not hide your single points of failure.

Every architecture has them. Most vendors hand over a tidy topology diagram and say nothing. We hand over a risk map that marks the single points — including the ones we have not solved — and you decide whether to pay for the fix.

Compliance

Where we actually stand

Three states, stated plainly: what we hold, what is underway, and what we do not offer. Anything else is a marketing claim.

Planned SOC 2 Type II

Program scoped; audit window typically 12 months.

Planned ISO 27001

Follows SOC 2 readout.

Not offered HIPAA

We implement HIPAA-aligned controls for customers who hold their own BAA. We don't sign as a covered entity.

Not offered PCI DSS

Out of scope as a service provider; we support your QSA evidence collection.

Book a 30-min cluster review

Bring your architecture diagram, or just the problem. We'll tell you what we'd do, what it would cost, and whether we're the right people for it.