Selected work

Systems in production, at scale.

Six engagements across fintech, energy, logistics, and health. Each one is a system we designed, built, and in most cases still help run. The names and figures below describe the kind of work we take on and the results we hold ourselves to.

The through line

Correctness under load, proven in production.

The projects here share a common shape. Real throughput, hard latency budgets, and a failure mode that costs money when it slips. We do not take on work we cannot instrument, and we do not call something done until it holds under traffic we can measure.

  • Measured, not assertedEvery headline number below came from production telemetry, not a benchmark run once on a laptop.
  • Owned end to endWe carried these systems from architecture through the first on-call rotation.
  • Still runningMost of these platforms are live today, several under our managed operations.

At a glance

The same standard across every sector.

Different industries, one way of working. Here is what each engagement demanded and the practice that led it.

Client Sector Practice Headline result
Nordbank Fintech Platform Overnight batch removed, 40 ms p99 settle
Helio Energy Energy Cloud Eleven clusters to one, 62% infra cost cut
Kestrel Logistics Logistics Data 3.5M events per second, 0.8 s end to end
Verta Health Health Applied AI Triage 71% faster behind guardrails
Fenix Payments Payments Applied AI Under 10 ms scoring, 40% fewer false positives
Meridian Freight Freight Cloud Detection cut from 22 minutes to under 3

Teams we have shipped production systems with since 2017

Nordbank Helio Energy Kestrel Logistics Verta Health Meridian Freight Aurora Mobility Fenix Payments Optic Retail
ECORTIQ took a ledger we were afraid to touch and turned it into the most reliable service we run. They understood the failure modes better than we did.
JK
Jonas Keller
VP Engineering, Nordbank

Across the portfolio

A track record measured in uptime.

42
Production systems delivered since 2017
99.98%
Median platform uptime over twelve months
3.5M/s
Peak event throughput handled in production
41 ms
Typical p99 API latency across services

Start

Ready to build the next one?

Tell us what you are building and where it hurts. We will tell you honestly whether we are the right team for it, and what the first release would look like.