Summary
I can make your webapp faster, more reliable and fault tolerant Backend engineer with 15+ years building and operating high‑load systems (Rails, Node.js, Python). Comfortable with event‑driven architectures, async I/O, Kafka/RabbitMQ, and PostgreSQL tuning. Known for rapid incident response (e.g., fixed a TimescaleDB crash loop within 4 hours by tracing it to an OS‑level cgroup memory cap in Kubernetes).
Experience
Fin tech — Ruby on Rails Developer
- Built and maintained services with Ruby on Rails (dry‑rb), Kafka, RabbitMQ, PostgreSQL, Kubernetes (k8s), Helm.
- Delivered 11k RPS on RoR + Postgres + Kafka; improved observability and stability.
Smartlead — AI Developer
- Shipped a WhatsApp chatbot for internal Q&A (Node.js, React, OpenAI API, LangGraph).
- Designed AI workflow orchestration; deployed on Kubernetes with queues for scaling.
Out of Office — Elixir Developer
- Travel app: Elixir, AWS RDS Postgres, Kubernetes.
- Reduced AWS RDS spend by ~30% via query/index tuning and container limits review.
Hill30 — Full‑stack / Systems Engineer
- Swingbyte — Golf training app + sync server (Android, RoR, PostgreSQL, S3).
- GemCMS — CMS (React, Elixir, PostgreSQL).
- NetAnalyzer — DPI analyzer (C/C++, Linux, .NET Core): filtered 99.9% of packets at C++ level; streamed ~0.1% into Kafka and processed with Spark; sustained 10 Gbit/s.
Scout — C# Developer
- ScoutSystem — GPS tracking (C#, MS SQL, high‑performance TCP).
Selected Projects & Wins
Fintech KYC Platform
- Split a monolith into microservices; Kafka event flows, JWT with Vault key rotation.
- Automated onboarding — manual setup time went from a full day to near zero.
AI Assistant (WhatsApp)
- Answered internal questions from docs; async workers for load spikes.
- Learned edge cases of async I/O under real traffic.
Infra Tools
- Go daemons for metrics/backups; optimized aiohttp server (~2M req/day).
- Added graceful shutdowns and retries to minimize deploy incidents.
High‑Load DPI Analyzer
- 10+ Gbit/s live traffic; kernel‑level filtering; streaming to Kafka/Spark.
- Fixed TimescaleDB crash loops in under 4 hours (root cause: cgroup memory limit).
- Cut AWS costs by ~30% via DB/infra tuning.
- Mentored juniors; two moved from QA to backend roles.