Diagnose
Map the owner, inputs, failure point, and first acceptance test before recommending scope.
Case Files — Systems Repair
Fixed-term repairs for broken workflows: bottleneck map, working fix, pass/fail test, and handoff package.
Every engagement starts with the failing flow and ends with named deliverables: map, fix, pass/fail test, owner notes, and handoff package.
A focused 2–4 week repair sprint for one broken handoff. We map the bottleneck, ship a working fix, run a pass/fail acceptance test, and leave the handoff package.
A lead form, inbox, or internal request queue is working enough to create work, but the owner, fallback, or follow-up rule fails silently. We map the handoff, repair the routing, and write the first acceptance test before widening scope.
Typically 2–4 weeks / one bottleneck / one owner handoff.
Form submits, no owner, no log, manual follow-up.
Routed to owner, logged, covered by fallback alert, and tested against a pass/fail path.
Map the owner, inputs, failure point, and first acceptance test before recommending scope.
Define the smallest working fix, data flow, handoff target, and pass/fail criteria.
Ship the scoped deliverable with visible milestones, instrumentation, and owner review points.
Run the acceptance test, package the runbook, train the owner, and hand over the keys.
Tools are chosen around the handoff: use what already works, add only what the owner can operate, and document the test path for anything new.
Owner, trigger, failing step, and the first test worth running.
A scoped repair in the real workflow, not a deck about the workflow.
Acceptance criteria written before the build is called finished.
Runbook, owner notes, credentials path, and the next safe change.
Input, owner, wait state, and failure mode.
What has to be true before the repair ships.
Notes, run path, and owner handoff in one place.
Final File
Send the page, form, prompt, process, or handoff that fails. We'll reply with the bottleneck we would test first and whether a Systems Sprint is enough.