Offshore delivery has a reputation problem, and it mostly earned it. The classic model rents you strangers from a bench: people who were working on someone else's product last month and will be on someone else's next quarter, managed by a vendor whose economics reward churn. Clients learn the same lesson at the same price point everywhere, and conclude offshore does not work.
What actually does not work is the bench. Our delivery centre in Vietnam runs on a different model, Assemble, Operate, Transition, and the difference starts before anyone writes code.
When a client commits to a sustained programme, we recruit the team around it. Not "who is free", but "who fits this stack, this domain and this client", hired through the same process we use for our own consultants, interviewed against the programme's actual needs. It takes longer than allocation from a bench, and the difference shows for years: the team's first loyalty is to the product they were hired to build.
This is also why the model has a floor. AOT is built for multi-quarter programmes. For a three-month project, recruiting a dedicated team makes no sense, and we say so rather than bend the model until it breaks.
The assembled team runs inside our delivery management: our engineering standards, our security controls, our code review discipline, with delivery managed from Sydney and one person accountable to the client for outcomes. The client works with the team daily, in their tools and ceremonies, but performance management, quality and continuity are our problem, not theirs.
That single point of accountability is the part clients notice. When something is late or wrong, there is no triangle of blame between client, vendor and individual. There is one phone number.
Here is the part that surprises people: when the team has proven itself, the client can take it over entirely. People, knowledge, codebase, the lot, transitioned into the client's own organisation on agreed terms.
Most outsourcing models treat client dependency as the revenue plan, which is precisely why clients distrust them. Building the exit into the model changes the incentives on both sides. We operate the team knowing it may one day be handed over, so documentation, standards and knowledge-sharing are maintained as if the handover were next quarter. The client invests in the relationship knowing it is not a trap. Some clients transition, many never do, and the option's existence improves the years before it is exercised either way.
A team that knows the product deeply because it has never worked on anything else. Engineers who turn up in the client's standups and stay for years, on Australian-overlapping hours that make collaboration normal rather than a relay race. Standards that match what our Australian consultants ship, because they are the same standards, checked by the same people. And a delivery centre that has put production systems live across financial services, health technology and property technology under exactly this arrangement.
The model is described in full on our About page, and the engagement structures it slots into are on the custom software practice page.
Ha Nguyen is Head of Delivery at Coder Trove's offshore delivery centre in Vietnam.