(206) 555-0128 ◇ Seattle, WA
kazuo.hirano@example.com ◇ linkedin.com/in/kazuo-hirano ◇ kazuohirano.com
Staff software engineer with 11 years in backend systems, currently leading the microservices platform for a travel booking company that runs 140 services and handles 22,000 requests per second at peak. Led the migration off a monolith that cut deployment lead time from 2 weeks to under 1 day, raised availability from 99.5% to 99.95%, and cut infrastructure cost 31%. Designs services with clear ownership and contracts so that 18 product teams can ship without waiting on each other.
Focus on Distributed Systems. Thesis on consistency trade-offs in event-sourced services.
Also holds AWS Certified Solutions Architect Professional, renewed in 2024.
- Led the migration of a booking monolith to 140 services over 3 years, cutting deployment lead time from 2 weeks to under 1 day.
- Raised platform availability from 99.5% to 99.95% by introducing SLOs, circuit breakers, and a Kafka outbox across 18 teams.
- Cut infrastructure cost 31%, about $2.6M a year, through Kubernetes right-sizing and removal of 22 duplicate services.
- Split a payments module into 9 services with saga-based transactions, supporting a rise from 4 to 31 payment methods.
- Cut failed payment captures from 0.8% to 0.1% by adding idempotency keys and retries with backoff to every service call.
- Introduced distributed tracing across 40 services, which cut mean time to resolve incidents from 3 hours to 35 minutes.
- Built 5 order management REST APIs in Java and Spring Boot handling 3 million orders a month at launch.
- Wrote the first contract tests for the order domain, catching 27 breaking API changes before release in year 1.
- Moved 12 batch jobs from cron hosts to containers, removing 8 servers and cutting nightly run time by 40%.
Booking Monolith Decomposition. Mapped the booking monolith into 11 bounded contexts with product and engineering leads, then used a strangler fig approach behind an API gateway to move traffic domain by domain over 3 years, which ended in 140 independently deployed services, deployment lead time under 1 day, and 0 customer-facing outages during cutover.
Golden Path Service Template. Built a service template with Spring Boot, OpenTelemetry, health checks, SLO dashboards, and an Argo CD pipeline, which cut the time to stand up a production-ready service from 3 weeks to 2 days and was used for 64 new services across 18 teams.
Multi-Region Failover. Designed active-passive failover for 30 tier-1 services across 2 AWS regions with replicated PostgreSQL and Kafka mirroring, tested through quarterly game days, which brought regional recovery time from about 4 hours to 12 minutes.
- Speaker at 3 regional cloud native conferences on migrating from a monolith without a freeze.
- Contributor to an open source Kafka outbox library, with 14 merged pull requests since 2022.
- Organiser of a Seattle distributed systems reading group that meets monthly with about 40 members.
- Chair the architecture review board that approves service boundaries and contracts for 18 teams.
- Mentor 5 senior engineers on the staff engineering track, 2 of whom have been promoted to staff.
- Incident commander rotation lead, running blameless reviews for about 30 incidents a year.

.webp)



.webp)


%20Which%20Should%20You%20Use.webp)