A HIPAA-adjacent patient platform moved off hand-built infrastructure onto a codified, observable AWS platform with predictable cost and tested disaster recovery.
(01) The challenge
Five years of organic growth left the platform with unknown dependencies, no infrastructure as code, and a monthly cloud bill rising faster than revenue. Backups had never been restore-tested.
(02) Our approach
We produced a complete architecture map, codified the estate in Terraform, right-sized workloads, moved stateful services to managed equivalents, and rehearsed disaster recovery end to end.
(03) The work
[Placeholder case study — all client details and metrics are illustrative and must be replaced with verified data before launch.]
The platform served clinics and patients, which made downtime both a business and a clinical concern. Yet the infrastructure had never been deliberately designed. It had grown: instances added during incidents, databases scaled by guesswork, and a backup process nobody had ever tested.
Two risks stood out:
Mapped everything. We traced dependencies across compute, storage, networking, and third-party services, then produced an architecture map and failure-domain analysis the team still uses today.
Codified the estate. 96% of infrastructure moved into Terraform modules with reviewed changes. Snowflake resources were eliminated or documented as exceptions with owners.
Right-sized and consolidated. Workloads were resized against actual utilisation, idle capacity was removed, and several self-managed services were replaced with managed equivalents that reduced both cost and operational load.
Rehearsed recovery. We ran a full disaster-recovery exercise, found it took over nine hours, fixed the gaps, and re-ran it until recovery completed in 22 minutes.
Cloud spend fell 41% in the first quarter while uptime improved from 99.4% to 99.99%. Disaster recovery is now a tested, documented, and regularly rehearsed procedure rather than a promise.
The cost reduction was a side effect. The primary outcome was an engineering team that could finally change its own infrastructure with confidence.
[Placeholder quote — replace with a real client testimonial before launch.] For the first time we have an accurate picture of our own infrastructure, and we can prove our recovery plan actually works.
(More work)
(Start here)
We'll map it to a measurable target and tell you honestly what it takes.
Book a call