Whether you book a stabilisation engagement or a single on-site bring-up day, the rhythm is similar: narrow the problem, make failure visible, fix with evidence, and hand back operable knowledge.
You send symptoms, MCU family, and hardware revision. We ask about power, temperature, and whether failures correlate with manufacturing lot. If the problem is purely mechanical or RF layout without firmware hooks, we say so and suggest partners.
Fixed fee or day-rate cap, deliverables, access requirements (repo, debugger, boards), and review cadence. NDAs executed if needed. Deposit invoice for fixed engagements.
Bench time in Walmer or remote debug. Logic analyser, SWO, or GPIO bit-banging logs—whatever makes the fault honest. No patch until we can turn it on and off.
Patches land in your VCS with conventional commits. You review diffs; we walk through ISR and power implications on a call. Second iteration included where scoped.
Soak tests, manufacturing witness, or OTA staging as agreed. Written report archived in your wiki. Optional retainer for the next hardware spin.