Common failure

  • Teams ship “happy path” flows without evidence trails.
  • Later, compliance asks for logs/approvals/controls → rewrites.
  • Ops starts doing manual work in spreadsheets → risk increases.

Approach

1) Convert compliance into product requirements

  • What data must be captured?
  • What actions require approval?
  • What evidence must be retrievable later?
  • What controls prevent misuse?

2) Build flows with “proof” built-in

  • Immutable event logs for critical transitions
  • Reason codes + attachments for manual actions
  • Maker-checker for high-risk operations
  • Search & export in backoffice

Design artifacts I create (as a PM)

  • Product Boundary Doc: what’s in MVP vs not
  • User Money Risk Doc: where money can break + safeguards
  • Compliance Flow Checklist: data capture, evidence, approvals

Practical examples (generic)

  • KYC state machine: pending → verified → rejected → re-review
  • Withdrawals: risk checks + velocity limits + escalation path
  • Manual adjustments: maker-checker + immutable audit entry
  • User restrictions: reason codes + time-bounded bans

Outcome

  • Fewer late-stage changes because evidence paths exist from day 1.
  • Ops can resolve issues without risky ad-hoc edits.
  • Audits become manageable because data is queryable and consistent.