I work with Go, distributed systems, and cloud infrastructure to turn messy backend problems into stable, measurable platforms.
The work I keep coming back to: backend services that survive real traffic, the latency and cost wins that move the SLA, and the visibility that lets a team debug its own systems — without me in the room.
Microservices that ship and stay shipped. gRPC + HTTP/2 between services, idempotent retries, circuit breakers, clean domain boundaries. The boring choices, taken seriously.
Profile-driven latency and infra-cost work. Measured in p99, throughput, and dollars — not in hand-waving. The wins are usually less code, not more.
OpenTelemetry across services with proper trace IDs, spans, and golden signals. Fewer noisy alerts, faster RCAs. The team finds problems before customers do.
Eight years across product startups and a high-traffic logistics platform. The arc: front-end → full-stack → cloud → backend → technical lead. Every move was a deliberate widening of scope.
Led FE on 20+ projects in React, SASS, Gulp. Mentored juniors into a delivery team.
stuck:Good code is the kind that survives a bad week.
Owned multi-tenancy onboarding service. Shipped Medicare-compliant middleware (AU).
stuck:Compliance is a forcing function for clean architecture — middleware is where systems actually live.
Cost APIs across AWS · Azure · OCI · GCP. Server migration tooling, ~40% time saved.
stuck:Every cloud is the same problem in a different uniform. Unit economics never lie.
Courier & assignment platform at ~25k req/s. ~50% latency cut, ~$2K/mo Redis reclaimed, 60× inference speedup.
stuck:The team is the system. RFCs and design reviews aren't overhead — they're the actual product.
Write-ups on backend engineering, performance work, and the observability that keeps systems honest. A few recent ones below.
A 30-minute call is the fastest way to figure out if there's a fit. Staff/Sr. role, advisory, or a second opinion on an architecture call. No deck, no pitch, just questions.