Kaya Retail Group
Eleven stores running on a legacy server with nightly CSV syncs and no single view of stock. We migrated them to AWS, rebuilt the integrations, and gave operations one live dashboard.
34%
Infra cost saved
99.9%
Uptime since launch
12 hrs
Admin saved weekly
The problem
Each store ran its own point-of-sale database, reconciled overnight by a CSV export that frequently failed silently. Nobody could answer how much stock existed across the group without phoning branches.
What we did
- Inventoried every workload, scheduled job and integration before touching anything.
- Defined the target environment in Terraform so staging genuinely matched production.
- Replaced the nightly CSV job with event-driven syncs and a dead-letter queue.
- Rehearsed the cutover twice, timed, with a documented rollback.
The outcome
The live cutover ran inside the scheduled window. Right-sizing during migration cut the monthly bill by about a third, and operations gained a dashboard that reflects stock within seconds rather than the next morning.
The migration was the calmest I have been through. They rehearsed the whole cutover twice, and the live window took eleven minutes.
Have a problem like this one?
Book a free 30-minute scoping call. Tell us the problem and we’ll tell you honestly what it takes to solve it.