Six phases to a platform you own
1. Foundations & security
Weeks 1-2Your own AWS environment stands up under MySamples' account from day one — networks, security, environments, and automated deployment, with every credential rotated onto a secure footing. You own the infrastructure outright from the first week.
What "done" means for this phase
- Staging environment deploys automatically from the main branch.
- No secret exists in code or configuration that is still valid.
- MySamples holds root access to its own AWS account; Wonderfiire and the developer operate on scoped IAM roles.
2. Core platform build
Weeks 2-5The heart of the rebuild: the doctor-facing ordering experience — catalogue, cart, checkout with per-product limits, and order history — plus a unified admin console bringing customers, orders, products, stock, and reporting into one place.
What "done" means for this phase
- A doctor can register, browse, order within limits, and see order history in staging, end to end.
- An administrator can manage products, customers, and orders and export every current report type from one console.
- Per-product limits are enforced server-side and cannot be bypassed from the browser.
3. Compliance & approval workflow
Weeks 5-7Live Ahpra verification for Australian doctors and HPI verification for NZ, a formal registration approval queue for your team, and an audit trail of every decision — only verified, approved practitioners can order.
What "done" means for this phase
- A registration with an invalid practitioner number cannot reach ordering, in either country.
- Every approval or rejection is attributable (who, when, on what evidence) in the audit log.
4. Business rules & automation
Weeks 7-9Suppliers become first-class with their own configurable claim limits, every ordering rule in the agreed register is built and demonstrated, the daily warehouse feeds ingest themselves on schedule with alerting, and full NZ localisation lands.
What "done" means for this phase
- The A3 register passes as an acceptance checklist: each rule demonstrated behaving identically to production (or better, where agreed).
- A supplier's limits can be changed by an admin without developer involvement.
- Feeds ingest for five consecutive days with zero human intervention; a deliberately-poisoned feed lands in the dead-letter queue and raises an alert within minutes.
5. Data migration & parallel run
Weeks 9-11A decade of doctors, orders, and products mapped and migrated — rehearsed on anonymised copies until it runs clean, reconciled down to individual records, then validated by your team in a structured UAT and parallel run.
What "done" means for this phase
- A full migration rehearsal completes with reconciled record counts and zero unexplained variances.
- The MySamples team signs off UAT on their real workflows.
6. Cutover, hardening & handover
Weeks 11-12DNS cutover in a low-traffic window with a tested rollback plan, production hardening with firewall, monitoring and alerts, a full documentation handover, and 30 days of hypercare while the new platform beds in.
What "done" means for this phase
- Production serves both domains from AWS with the legacy site fully retired from the request path.
- The documentation pack is delivered and walked through with the team.
Build updates
Updates will appear here as the build gets underway — expect the first one within days of kickoff.
What's coming
- A one-command development environment replicating the full platform stack
- A built-in health check reporting whether the application, database, and job queue are up
- All application settings centralised into one documented configuration layer
Questions?
Anything on this page you'd like walked through — email Ben or call any time.