The handoff: from our delivery pod to your own team.

Shadow, Pair, Lead, and Observe. In Shadow, Silicon Prime builds while your engineers read along, join planning, and observe reviews without touching production. In Pair, both teams build together to transfer tacit knowledge. In Lead, your team owns tickets, releases, and decisions while Silicon Prime reviews. In Observe, you run it and Silicon Prime stays on call for a brief, predefined period—covering real-world issues without creating long-term dependency.

Most handoffs fail because they're abrupt—a vendor works in isolation, then leaves with only documentation. Silicon Prime integrates the transfer into the work itself so ownership moves continuously rather than in a single overwhelming week. The crossover point is planned in advance, so both teams know exactly when your engineers assume full ownership. The slope avoids the knowledge cliff that strands an internal team after a vendor departs.

Not in the Shadow phase—there your engineers only gain context by reading along, joining planning, and observing reviews while answers are still straightforward. Your team begins contributing to the codebase in the Pair phase, building alongside Silicon Prime. Full ownership of tickets, releases, and critical decisions comes in the Lead phase, where the first independent release by your team marks a significant milestone.

By what survives them. The stated rule is that if the program requires the pod's return after three months, the handoff was not successful. The aim is that ownership transfers cleanly and the program thrives independently after departure. As the post puts it, a pod unable to depart cleanly was never a true pod—it was a dependency.

Because the most valuable knowledge is tacit and isn't easily documented. The post states the handoff is not a document you write at the end—it is the thing you were doing the whole time. The Pair phase exists specifically to transfer that tacit knowledge by building together, which a closing document can't capture. Continuous transfer is what lets the slope work instead of a cliff.

You run the system, and Silicon Prime remains on call for a brief, predefined period. This window covers potential issues that surface under real-world conditions once your team is fully operating it—without establishing any long-term dependency. It's a deliberately time-boxed safety net, not an open-ended retainer, consistent with the principle of measuring success by what survives the pod's departure.

Stage your engineers to match the phases: assign people to read along and join planning during Shadow, free up capacity to build alongside the pod during Pair, and prepare them to own tickets, releases, and decisions by the Lead phase. Because the crossover point is planned in advance, you can schedule when internal owners assume full responsibility and reserve light involvement for the predefined Observe window.