tempo
PublicDistributed cron with exactly-once delivery, per-tenant rate limits and a Raft-backed schedule log.
interface Engineer { name: string role: string stack: string[] } const kaito: Engineer = { name: "Kaito Ishida", role: "Senior Software Engineer", based: "Kyoto, Japan", stack: ["TypeScript", "Go", "Rust", "Postgres"], focus: ["distributed systems", "developer tooling"], years: 11, } // I build reliable back-ends and the tools that ship them
Public commits, reviews and issues across every repository over the last year. Quiet weeks are real too — sustainable pace beats heroics.
A slice of the open-source work — infrastructure I reach for on real systems, shipped with docs, tests and a changelog.
Distributed cron with exactly-once delivery, per-tenant rate limits and a Raft-backed schedule log.
Embeddable LSM key-value store with a crash-safe write-ahead log — 1.2M writes/sec on a laptop.
Virtualized React data-grid that renders a million rows at a steady 60fps with keyboard-first navigation.
Structured logging and tracing SDK with zero-allocation hot paths and drop-in OpenTelemetry export.
Migration linter that blocks unsafe DDL before it reaches production — catches locks, rewrites and drops in CI.
Interactive EXPLAIN ANALYZE visualizer for untangling Postgres query plans, node by node, in the browser.
Opinionated but not dogmatic. Boring where it counts, sharp where it pays off.
I'm a software engineer who has spent most of his career on the unglamorous parts of software — schedulers, storage engines, and the pipelines that stay up while everything around them changes.
I care about correctness under load, honest failure modes, and tooling that makes a team measurably faster. These days I lead a small platform group in Kyoto and write open source in the evenings. I read the papers, I take the pager, and I still enjoy deleting more code than I add.
Lead the data-platform group. Designed a multi-region event bus moving 12B events/day at four-nines availability.
Rebuilt the billing pipeline end to end; cut reconciliation errors 98% and halved month-end close time.
Owned storage and queueing for a platform serving 4M daily users across the region.
First role. Shipped the public API that became the company's core product.
Graduated top of cohort; thesis on consensus under partial failure.
Best for staff and principal roles, fractional platform leadership, or a second opinion on an architecture you're not sure about.
kaito@kyoto:~$ cat contact.txt
Status available — open to staff / principal roles
Timezone JST (UTC+9) · async-friendly, remote-first
Reply within one business day
PGP 4096R / 0xA1F3 9C22 7E10 DD84
kaito@kyoto:~$ ./say-hello