01 · Fintech and payments

Money moves on a schedule you don't control.

Payments platforms carry the strictest combination we see: PCI DSS scope, a ledger that cannot be wrong, and an on-call rotation that has to work at three in the morning. The platform is usually three clusters and five years of decisions, and everyone already knows which parts are fragile.

We work on the parts that fail loudly or audit badly: the SLOs nobody wrote down, the cluster that costs more than it should, and the change that cannot be traced to an approval.

  • SLOs and error budgets for the payment path
  • PCI DSS evidence from the platform, not a policy document
  • Multi-cluster cost review and rightsizing
  • On-call design that survives one engineer being asleep

Reliability & Cost Audit

02 · Healthtech

Clinicians notice a slow platform before a dashboard does.

HIPAA sets the floor. The real requirement is that clinical staff can work, and that the evidence for the audit already exists when the auditor asks for it. Most healthtech platforms we see have good intentions and three-day log retention.

We implement the controls in Terraform and Kubernetes, so access, logging, encryption, and recovery produce their own evidence instead of a document claiming they exist.

  • HIPAA controls as code, with the evidence attached
  • Audit logging and twelve-month retention
  • Restore drills that are timed and written down
  • Access review tied to your identity provider

Audit-Ready Platform

03 · Insurance and banking

The auditor arrives with a list, and the list is technical.

SOC 2, E&O reviews, and internal audit all end in the same place: can you show what changed, who approved it, and what happened when it broke? Most teams can answer that after a week of digging through Git and a chat history. We make it a query.

The same work usually pays for itself before the audit: cost review on the legacy Azure estate, drift eliminated, and an on-call rotation that does not depend on one person.

  • Change traceability from pull request to deploy
  • Evidence export an auditor can read
  • Disaster recovery tested, with a written result
  • Legacy cloud cost review

Fractional SRE

04 · B2B SaaS

Big enough to need an SRE, too small to hire one.

Twenty to two hundred engineers, Kubernetes on AKS or EKS, and an enterprise customer asking for a SOC 2 report. The platform is not on fire. It is just unowned, and every quarter it gets a little more expensive and a little harder to change.

This is the company Arcline was built for: the audit first, because it tends to pay for itself, then a retainer that keeps the practice in place.

  • Reliability and cost audit, one week, read-only
  • Fractional SRE, ten to fifteen hours a week
  • GitOps and drift control
  • SLOs agreed with product, not just engineering

Fractional SRE

Not on this list?

We would rather say no than stretch.

Most of our work is one of those four. If you are somewhere else — a marketplace, a logistics platform, a public-sector system — say so anyway. We will tell you honestly whether we are the right fit, and point you somewhere better if we are not.

Which one is your platform?

Tell us the industry, the stack, and what is coming up. We will tell you where the audit would start.