Why this matters
- Guarded rollouts with live SLO checks (latency/error).
- Auto-rollback < 2 minutes when error budget burns.
- Observability built-in: traces, metrics, logs, dashboards.
I used AI tools for tests/docs only; core logic is hand-written and benchmarked. See
AI_USAGE.md.
Featured Projects
1) Canary Platform — Feature Flags + Auto-Canary with SLO Guardrails
- REST API + .NET SDK; iOS sample client with offline fallback
- OpenTelemetry dashboards (p50/p95/p99), error rate, RPS
- Reliability: health checks, retries with jitter, circuit breaker, chaos drill & postmortem
- Numbers: rollback median 1m47s, p95 latency −18% after cache
Links:
- Repo: https://github.com/Landers99/canary-platform
- Demo: (coming Friday)
- Case study: todo
2) redis-lite (BYOX) — protocol + AOF + LRU + mini YCSB
- RESP parser, GET/SET, persistence, eviction
- Bench graphs vs baseline; property tests
Links: https://github.com/Landers99/redis-lite
Demos & Posts
- 🎥 Demo #1: Canary rollback under SLO breach (3:12)
- 📝 Blog #1: SLO-guarded rollouts that save your error budget
- 🎥 Demo #2 (coming): iOS offline toggle + background refresh
About & Contact
Sammamish, WA · Open to backend, platform, and iOS roles
Email: smithlandrey@gmail.com · LinkedIn: https://www.linkedin.com/in/jonathan-l-smith/