Champion Health

Champion Health’s insurance platform manages a hierarchy of 100+ agents across multiple broker organizations. When an agent needed to view data from another branch — for collaboration, oversight, or compliance — there was no formal process. Requests came in via email or Slack with no tracking, no audit trail, and no way for the target agent to approve or object.

The team knew they needed a self-service workflow with peer approval, automatic escalation, and broker-boundary enforcement — but had no PRD, no technical spec, and no prototype. They needed someone who could define the problem, design the system, and prove it works — fast.

Started by writing a comprehensive PRD: database schema, API contracts, state machine spec, email notification matrix, admin dashboard requirements, seed data, and acceptance criteria. Then built the entire system — a seven-state request lifecycle (PENDING → OBJECTED → ESCALATED → APPROVED / OVERRIDDEN / DENIED / CANCELLED), branded SES email notifications with deep-link approve/object buttons, and configurable auto-escalation with halfway reminders.

Deployed a live prototype on GCP so the team could interact with the full workflow within the week. The admin dashboard includes status tabs, search, date-range filters, sortable columns, and stale-request indicators. Every state transition is logged in a full audit trail. Broker-boundary enforcement ensures agents can only request visibility within their own organization. Approved requests automatically create grants in the existing system.

In one week the team went from an undefined problem to a live prototype running on GCP — complete with a PRD the engineering team could build against, seed data modeling their real agent hierarchy, and a working demo they could click through. The prototype validated the state machine design, the email notification flow, and the admin dashboard UX before committing to a full production build.