Platform migration to Kubernetes
Moved 40+ services from hand-managed VMs to EKS with GitOps. Every environment is now a pull request away.
40+services moved
0downtime min
DevOps · Platform · SRE
Andrei
Andrei
Shilkov
I build platforms where shipping is boring — fast pipelines, self-healing clusters and infrastructure as code, so teams deploy on Friday without fear.
❯ whoami
andrei — devops / platform / sre
pipeline · main
run #4812 · running · total 0.0s
commit
—
build
—
test
—
scan
—
deploy
—
verify
—
/metrics
Reliability you can graph
What the platforms I run look like in numbers: shipping more often, recovering faster, spending less.
Uptime · 12 mo
99.98%
SLO 99.9 · error budget 80% left
Deploys / month
74
▲ 4.1× vs. last year
MTTR
12min
▼ from 2h 40m
Cloud cost
−38%
spot + rightsizing + autoscale
Deployments per month
prod · Oct ’25 – Sep ’26API latency p95
live118ms
p95SLO 250 ms
Rolling update
48/48 readyv2.14→ v2.15
oldpendingnew
/stack
Tools I reach for
Opinionated, boring, well-understood — picked to be operated at 3 a.m.
Cloud
AWSGCPHetznerCloudflare
Orchestration
KubernetesHelmKustomizeDockerKarpenter
IaC & GitOps
TerraformOpenTofuAnsibleArgo CDFlux
CI/CD
GitHub ActionsGitLab CIJenkins
Observability
PrometheusGrafanaLokiOpenTelemetry
Security
VaultTrivyOPA / KyvernoIAM
Data
PostgreSQLRedisKafka
Code
GoPythonBashLinux
/certifications
Certified, and verifiable
Click a badge to check the credential.
/case-studies
Selected work
Problem, what I changed, what it moved.
CI pipeline from 25 to 6 minutes
Layer caching, parallel test shards and ephemeral runners. Developers stopped context-switching while waiting.
−76%build time
4×deploy rate
Observability & cost control
SLO-based alerting instead of noise, plus spot nodes and rightsizing driven by real usage data.
−38%cloud bill
−70%pager alerts
/experience
Where I've shipped
2023 — now
Senior DevOps Engineer · Fintech scale-up
Own the Kubernetes platform, CI/CD and observability for 60 engineers. Led the move to GitOps and SLO-based on-call.
2020 — 2023
DevOps Engineer · E-commerce platform
Terraform for all infrastructure, zero-downtime deploys, and Black Friday capacity planning.
2017 — 2020
Linux System Administrator · IT integrator
Automated fleet configuration with Ansible and built the first CI pipelines for the dev team.
/contact
Let's make your