Services
A team that stays on one product.
Two to six engineers with a named technical lead, working in your repositories on your cadence for the length of the engagement. The team is not shared with another client, and the engineers who start are the engineers who finish.
How a dedicated team runs
Who you get
Median engineer experience across the 21 people here is 9 years and median tenure with us is 2.8 years, so a four-person team is normally four people who have each shipped production systems for the better part of a decade. One of them is the named technical lead and runs review, estimation, and escalation on our side.
We staff against ten technologies: Java/Spring, Go, TypeScript, React, Python, PostgreSQL, Kafka, AWS, Kubernetes, and Terraform. English is B2+ for engineers and C1 for team leads, assessed in the first call rather than from a CV. There is no bench: the 21 engineers here are on 6 active client engagements, so a team of four is assembled from people finishing other work, and the start date reflects that instead of hiding it.
Ramp-up takes 10–15 business days
From signature to a first merged change is 10–15 business days: repository and environment access, a walk through the domain with your engineers, and one small change shipped through your own review process before the team takes a full sprint's load. Those days sit inside the three-month minimum, not on top of it. If your codebase needs more than fifteen days of context before a first change is safe, say so on the first call: it usually means starting with two engineers rather than five.
Hours, overlap, and who owns the backlog
Priority stays with your product owner. We do not run a parallel backlog, and we do not report progress in a format you have to learn. Overlap with US East coast hours is 5–7 hours a day, with US West 2–4, and EU clients get a full working day. Hours, review, and escalation are written into the working agreement before anyone writes code.
When an engineer leaves
Annual attrition is 11%, so across a 14-month engagement this is an event to plan for rather than a risk to describe. A replacement joins three weeks before the leaver's last day and takes the work over during that overlap. Either side can end the engagement on 30 days' notice; the three-month minimum is the only lock-in in the contract.
Where this has run
Two dedicated-team engagements are written up in full, including the one that ended early. Team sizes were four and three engineers; durations were 14 and 19 months.
All four case studiesDispatch platform rebuild
Routing pulled out of a Rails monolith into a Go service one route type at a time, with the monolith serving live traffic throughout. On the routes moved, p95 dispatch latency went from 2.4 s to 780 ms.
Claims ingestion pipeline
A nightly batch replaced with Kafka-based ingestion and a daily reconciliation report; dropped records went from about 40 a week to zero across 11 months. The client was then acquired, moved onto the acquirer's platform, and the engagement closed early.
Two to six engineers, merging inside three weeks.
A named technical lead on our side, your product owner on yours. If an engineer leaves, the replacement overlaps them for three weeks and you are invoiced for one seat.