Restaurant tech · Payments Senior Software Engineer 2024 — 2026 Tokyo, Japan (onsite)
Systems that survive the lunch rush.
Restaurant ordering and payment platforms live or die at noon. At Quad in Tokyo I built and ran services for ordering, payments, health food, and marketing campaigns — systems where traffic arrives in daily spikes and a failed order is a lost customer.
What made it hard
Traffic concentrates into meal-time spikes — capacity is planned for the peak, not the average
Payments must stay correct under load; "mostly worked" is just a refund queue
Onsite in a Japanese engineering team, where precision in communication counts as much as precision in code
The flight plan
Built high-performance APIs in Python (FastAPI) and Go behind the ordering and payment flows
Ran workloads on GCP Cloud Run, with Memcache-backed caching where latency mattered most
Managed infrastructure as code with Terraform, keeping every environment reproducible
Shipped responsive React frontends for customer and campaign surfaces
What held
- Services tuned for the hour that matters — the spike, not the demo
- Infrastructure changes became reviewable pull requests instead of tribal knowledge
- Fifteen months delivering onsite in Tokyo — proof I hold up inside a high-standard, non-English-first engineering culture
Debrief
PythonFastAPIGoReactGCPCloud RunTerraformMemcache
Cache-asideAutoscaling on managed runtimesInfrastructure as codeAPI-first services
Details are anonymized to respect confidentiality — the patterns are real, the internals stay private. Happy to go deeper on the architecture in a call.
Have a mission like this?
One freelance slot per quarter — the same rigor, on your product.