Cases that sell the work

Three outcomes from production environments — AI platforms, banking data pipelines, and large-scale SaaS SRE.

01 AI · Platform

SPIMEX AI chatbot — production platform in hours, not days

Challenge: Launch a regional, fault-tolerant Kubernetes footprint on VK Cloud for an AI chatbot product, with a single clear release path for many services — without multi-day manual infrastructure work.

What I delivered: Designed and operated Kubernetes across 3 availability zones. Codified provisioning with Terraform, Helm, and GitLab CI so infra moved from multi-day manual setup to pipeline-driven deploys in hours. Built one unified release pipeline for 10+ services (monitoring, ingress, messaging, PostgreSQL, backend) and eliminated manual deploys across environments. As sole infrastructure owner, presented architecture and release plans to client stakeholders and shipped the production platform on schedule.

VK Cloud Kubernetes Terraform Helm GitLab CI PostgreSQL

02 Banking · Data

Halyk Bank — Oracle to Tarantool CDC at 22× throughput

Challenge: Stand up a cloud banking environment and prove an Oracle → Tarantool data path under real load — OpenShift, Oracle 19c, and enterprise CDC tooling in one coherent pipeline.

What I delivered: Deployed a VK Cloud stand with OpenShift 4.21 and Oracle 19c / OpenLogReplicator, then delivered the full pipeline: Oracle → OLR → Tarantool CDC → TQE. Ran a successful load test that scaled from 50 to 1,100 TPS (22× throughput) over ~5 hours 20 minutes — giving the bank measured confidence in production readiness.

OpenShift Oracle 19c OpenLogReplicator Tarantool CDC TQE VK Cloud

03 SaaS · SRE

Mindbox — 15+ clusters, cheaper Kafka, cleaner alerts

Challenge: Run production infrastructure for a marketing automation SaaS at scale — many Kubernetes clusters, noisy alerting, costly Kafka, and too much manual release work.

What I delivered: Operated 15+ production Kubernetes clusters with a full observability stack (Grafana, Prometheus, Thanos, Alertmanager, Graylog, Strimzi). Standardized 85–90% of company alerts; moved releases, MySQL backups, and chaos tests into GitLab, GitHub, and Octopus Deploy. Codified cloud and bare metal in Terraform/Ansible. Migrated Apache Kafka to Kubernetes: 20–30% lower Kafka spend, higher reliability, and workload autoscaling. Also deployed HA Kea DHCP for 500+ servers and led on-call for SLA/SLO incidents.

Kubernetes Kafka / Strimzi Prometheus Thanos Terraform Ansible GitLab CI

Want the same outcomes on your team?

Telegram, email, or LinkedIn — pick what is fastest for you. I am open to remote full-time roles and relocation worldwide.

Contact