SOKOL CLOUD

Infrastructure engineering, hands‑on.

More than eight years of hands‑on work on infrastructure for fintech, banking and high‑load products. We design the architecture and write the code ourselves, from network routing to the developer platform.

Software development

Alongside infrastructure, we write software ourselves, from network clients to cloud tooling. We take it from idea to production: architecture, code, CI/CD, hosting and support.

Open solutions for everyone

Products and open‑source tools that anyone can install and use, like our VPN app proxyapp and the Rustket desktop client.

Custom solutions for one client

Software built around a specific problem in your business: integrations, internal tools, automation and services that plug into your infrastructure.

Go, Rust, Python, Flutter, Android

Areas of expertise

Cloud and hybrid architecture

Fault‑tolerant environments on AWS and GCP, and hybrid setups that join public cloud with bare‑metal data centres. Dev, stage and prod are built the same way.

AWS, GCP, OVHcloud, Proxmox, Transit Gateway, shared VPC, Route 53

Networking

Dynamic routing between clouds and data centres, site‑to‑site tunnels, NAT and DNS. Network changes rolled out without downtime.

BGP, BFD, ECMP, BIRD, IPsec, Calico, Istio, NetBird, Cisco

Kubernetes

Production clusters on EKS and GKE: autoscaling, resource requests sized from real usage, disruption budgets, policy‑as‑code and GitOps.

EKS, GKE, Helm, Werf, ArgoCD, HPA, Cluster Autoscaler, KEDA, Kyverno

Infrastructure as code

Everything is described in code, reviewed and applied from a pipeline. When a provider is missing, we write one.

Terraform, Terragrunt, Ansible, custom Terraform providers in Go

CI/CD and developer platform

Pipeline templates and golden images that let product teams ship services on their own. Package registries for PHP, JS, .NET and Java.

GitLab CI, Werf, Helm, Docker, ProGet

Observability

Metrics, logs and alerts that point to the cause, not just the symptom. Alerts go straight to the service owner.

Prometheus, VictoriaMetrics, Grafana, ELK

Data services

Running the stateful parts: message queues, databases, caches and object storage, including replication between sites.

Kafka, PostgreSQL, Redis, Valkey, MinIO

Security and compliance

Secrets management with automatic key rotation, single sign‑on, and infrastructure that passes audits in regulated finance.

HashiCorp Vault, External Secrets, OpenID Connect, Azure AD

FinOps

Cloud spend tied to real usage: rightsizing, Savings Plans and Spot, and removing what nobody uses.

AWS Savings Plans, Spot, cost dashboards

Results from past projects

  • Time to deploy a new service cut from 10 days to 8 hours with pipeline templates and reusable Terraform modules.

    Fintech platform

  • 99.99% uptime for critical payment and banking systems.

    Fintech platform, digital bank

  • Infrastructure incidents down 80%, recovery time (MTTR) down 90%.

    Digital bank

  • Cloud costs down 30% with no loss of performance.

    Fintech platform, digital bank

  • Infrastructure scaled through 300% year‑over‑year customer growth without interruptions.

    Digital bank

How we work

Code, not clicks

Every change to production goes through code review and a pipeline. No manual edits.

Measure first

Proposals start from data: real usage, real costs, real error rates.

Pilot, then roll out

New setups run on one node or a test environment before they reach production.

Agreed changes only

Production work is announced and approved. If the plan changes, we stop and get approval again.

Written down

A runbook for every system and a root‑cause review for every incident.

Clear ownership

Every service has an owner, and its alerts and costs go to that owner.

Have a project in mind?

Tell us about your infrastructure and what you want to change.