He will not let us use his name. That is fine. The system he rebuilt routes roughly one in every nine parcels moving across Southeast Asia. He has been doing it for eleven years, through three CEOs and one near-bankruptcy.
What follows is a long conversation, edited for clarity, about what it actually takes to keep a routing engine alive across a decade of regulatory shifts, currency shocks, and at least one weekend where a single misconfigured cache lost the company eight million dollars in COD reconciliation.
On the rewrites
Three of the five biggest rewrites were triggered by regulatory shifts, not by technical pressure. The system itself was fine — the law changed underneath it. The team kept an explicit 'postponed decisions' list and revisited it monthly. That list is, in his words, the single most important artifact the engineering organization owns.
"We don't measure success by uptime anymore. We measure it by how few of last quarter's emergencies showed up again this quarter."
The real story is not the architecture. It is that the same person stayed in the room for eleven years and held the institutional memory. Most of the failures he has seen in peer companies, he says, came down to that person leaving and nobody noticing for six months.
Dispatch Newsletter
One operator-grade breakdown every Tuesday. SEA tech, AI economics, execution frameworks. No fluff.
