10h → 20min
Automated the entire infrastructure deployment process for edge devices (mini-PCs)
Mission KS · DevOps · Remote or relocation
Kalislav Smirnov
DevOps Engineer Aerospace Engineer
I build the platforms that banks and government agencies trust with their releases.
4.5 years of Kubernetes, GitLab CI/CD, Ansible and Terraform, most recently in fully air-gapped networks, with security checks built into every pipeline. Spacecraft engineer by degree. Based in Tbilisi. Open to remote work for a company in any country or to relocation to the US, the EU, Cyprus or Kazakhstan.
Open to offers now
Remote · any country
Relocation · US, EU, Cyprus, Kazakhstan
Georgia, Tbilisi
- 5 bank & government platforms
- 100+ VMs across datacenters
- 2w → 3d release cycle
- 100% of pipelines scanned
- 10h → 20min device setup
- ● Open to offers now
Open source, live from GitHub
01 Impact
5
End-to-end DevSecOps platform implementations for banking and government clients
100+ VMs
Multi-datacenter Kubernetes platforms with disaster recovery protocols
2w → 3d
Release cycle for 20+ microservices
Live from the homelab this site runs on how it works ↓
02 Payload
Systems that banks trust with their releases.
Client names are left out on purpose. The numbers are from the projects themselves.
Federal company · 90+ development teams · 200+ products · team of 3 engineers · 1 year
Secure development process across the whole company
Before
Dozens of teams shipped code with no shared security checks, and open source packages reached developers unchecked.
What I did
- Deployed OSA Firewall and put it into the package supply chain, so open source dependencies are checked before developers get them.
- Deployed an ASOC platform that collects findings from every scanner in one place.
- Added Gitleaks, Trivy, SCA and OSA checks to the shared GitLab CI templates, so teams got them without writing pipelines by hand.
- Trained DevOps engineers and developers to work with the findings. Together with the client we prepared and agreed 30+ process documents and instructions.
Result
100% of pipelines run security checks
Major bank · air-gapped network
Signed package delivery into a closed network
Before
The bank's Nexus repository has no internet access, and every package that enters it has to come from a trusted source.
What I did
- Built a delivery service in Python that runs on Kubernetes.
- Every package is signed with GPG, and the signature is checked before it lands in Nexus.
Result
Developers in the closed network get the packages they need, and each one can be traced to a signature.
Investment bank · edge devices
Mini-PC setup automated end to end
Before
Each edge device (a mini-PC) was set up by hand, which took a full working day.
What I did
- Automated the entire infrastructure deployment with Ansible and custom scripts.
Result
10h → 20min to set up one device
Investment bank · core application
From a monolith to 10 services on Kubernetes
Before
The core application was a single monolith, and there was no monitoring for a service based setup.
What I did
- Designed and ran the first migration of the monolith to 10 containerized microservices on Kubernetes.
- Wrote Helm charts for all 10 services.
- Built monitoring and alerting from scratch with Prometheus, Grafana and ELK.
Result
Any service deploys or rolls back with one command, and the team sees problems in real time.
03 Flight log
Four and a half years of taking manual work out of releases.
-
DevSecOps Engineer
Feb 2025 - Aug 2026Swordfish Security Full-time
Deploying and maintaining DevSecOps platforms for enterprise clients in banking, fintech, and government sectors, including fully air-gapped environments.
- Led 5 end-to-end DevSecOps platform implementations for banking and government clients, from architecture design to final handover.
- Engineered and deployed a secure, GPG-signed package delivery service using Python and Kubernetes for a major bank's air-gapped Nexus repository.
- Automated the integration of SAST, DAST, and SCA tools (Semgrep, Trivy) into GitLab CI, enabling security scans on 100% of new code commits.
- Developed a suite of 15+ reusable Ansible roles to standardize deployments across client environments, reducing setup time for new projects by an estimated 30%.
- Designed and managed multi-datacenter Kubernetes platforms (100+ VMs) with disaster recovery protocols for high-availability financial systems.
- Prepared and conducted product demonstrations and acceptance testing (Nexus, GitLab, DevSecOps products).
- Maintained internal infrastructure: GitLab, proprietary products, RnD and PROD environments
-
DevOps Engineer
Oct 2024 - Feb 2025Sinara Investment Bank
- Automated the entire infrastructure deployment process for edge devices (mini-PCs), slashing setup time from 10 hours to 20 minutes using Ansible and custom scripts.
- Architected and executed the initial migration of a core monolithic application to 10 containerized microservices on Kubernetes.
- Built a complete observability stack from scratch using Prometheus, Grafana, and ELK, providing real-time monitoring and alerting for the new microservices architecture.
- Developed and managed Helm charts for all 10 services, enabling one-command deployments and rollbacks.
-
DevOps Engineer
Apr 2022 - Oct 2024Engineering & Manufacturing Company
- Overhauled the CI/CD process by implementing automated testing and deployment pipelines in GitLab CI, cutting the release cycle for 20+ microservices from 2 weeks to 3 days.
- Automated manual deployment checklists with Ansible and Bash scripts, reducing hands-on deployment time from over 4 hours to under 30 minutes per release.
- Led the containerization of 20+ legacy Java microservices using Docker, standardizing environments and eliminating configuration drift between staging and production.
- Managed and maintained the company's production infrastructure, consisting of 50+ Linux servers across two data centers.
Skills
Platform
Automation & IaC
Cloud
CI/CD
Security
Observability
Data & messaging
Engineering
04 Origin
Trained to build spacecraft. Now I build infrastructure.
I studied spacecraft engineering at Bauman Moscow State Technical University from 2020 to 2026, alongside my DevOps work.
Thesis · graded excellent
Suborbital spaceplane for intercontinental commercial flights
Integrated Master's degree (Specialist, Engineer): a 6-year programme, 360 ECTS, equivalent to a Bachelor's plus a Master's.
Design, Production and Operation of Rockets and Rocket-Space Systems. Specialization: crewed and uncrewed spacecraft and systems.
What I did
- 3D models and design drawings of spacecraft structures and assemblies.
- Strength, vibration and dynamics calculations for spacecraft structures.
- Aerodynamics and gas dynamics, flight theory and ballistics, propulsion systems.
- Modelling of automatic control and spacecraft stabilization.
- Research work and a research-and-production internship, both graded excellent.
Learning & education
Kubernetes certifications: CKA, then CKAD and CKS
In progressPreparing for the Certified Kubernetes Administrator exam first, then Certified Kubernetes Application Developer and Certified Kubernetes Security Specialist. Started in 2026.
Follow the preparation live →Integrated Master's degree in Aerospace Engineering (Specialist, Engineer)
Bauman Moscow State Technical University, 2020 - 2026
English - B2
Russian
05 Ground control
Not slides. This site runs on my own servers, live.
Homelab
This site runs on a mini-PC at my home.
Architecture
Built from the Terraform code in the homelab repo. terraform/outputs.tf
- Proxmox VE on ZFS.
- Local DNS with ad blocking and wildcard Let's Encrypt certificates issued via DNS-01.
- No open ports. I connect over Tailscale, and this site is published through Cloudflare Tunnel.
- Monitoring with Gatus (configured in YAML) and Uptime Kuma.
Recent changes in the repo
All commitsProjects
Public repositories on GitHub, loaded live.
Live sites
Kubernetes node size calculator: allocatable CPU/memory and max pods for every EKS, GKE and AKS instance type.
Tapeclip — Mac app that turns a track into a music video.
Sandbox
A real shell in a throwaway VM on my Proxmox host.
- Isolated network: no internet and no access to other hosts.
- 1 vCPU capped at 50%, 768 MB RAM, 6 GB disk.
- Rolled back to a clean snapshot every 30 minutes.
- Up to 5 people at a time.
guest@sandbox:~$
Latest field notes
All field notes →
06 Contact
Ready for the next mission.
Remote for a company in any country, or relocation to the US, the EU, Cyprus or Kazakhstan. Open to offers now.
For recruiters
Where are you based, and will you relocate?
I live in Tbilisi, Georgia (GMT+4). I am open to remote work for a company in any country and to relocation to the US, the EU (Cyprus included), Kazakhstan or another popular destination.
What type of employment are you looking for?
Full-time work as an employee, on a B2B basis or on a contract.
When can you start?
Right away. I am open to offers now.
Do you need a visa?
Remote work from Tbilisi needs no visa. For relocation I need a work permit, for example the EU Blue Card or a US work visa.
What roles are the best fit?
DevOps and platform engineering: Kubernetes, GitLab CI/CD, Ansible and Terraform, with security checks built into pipelines. I have the most experience with banks, fintech and government projects, including air-gapped environments.
What languages do you work in?
English (B2) and Russian.