Skip to content
lastdevops

Kubernetes consulting and support

We've worked on container orchestration since 2014, when our co-founder started building Last.Backend, his own orchestrator. Today we design, migrate and operate Kubernetes on managed clouds, bare metal and air-gapped sites.

Talk to an engineerStart with an audit

An engineer replies, not a salesperson — within one business day.

  • Container platforms since 2014
  • Zero-downtime migration
  • Air-gapped and GPU clusters
  • Co-founders do the work

Official cloud partners

  • Yandex Cloud
  • Selectel
  • Timeweb Cloud
among our clients
  • Novotek
  • RZD
  • SberCorus
  • Otus
  • Besttrip
  • JedAI
  • IMSystems
  • Builddocs
  • Headframe
  • RTDS
Case studies →
symptoms.log

When teams come to us

Upgrades keep getting postponed

Nobody is sure what will break, so the cluster falls further behind supported versions.

Costs drifted from usage

Resource requests and node pools no longer match what the workloads actually need.

Every team deploys differently

Helm here, raw YAML there, somebody's script in between.

Is Kubernetes even right for you?

You want an honest answer before investing in a platform you may not need.

scope.md

What we do

Cluster design

Managed or self-managed, on-premises and air-gapped, sized to your workloads and multi-zone by default.

Migration

From VMs, Docker Compose or another platform, one service at a time, without downtime.

GitOps delivery

ArgoCD and Helm, with environments described in Git and drift visible immediately.

Upgrades and lifecycle

Tested in staging first, with a written runbook and a rollback plan, in a window you choose.

Security baseline

RBAC, network policies, secrets in Vault and image scanning in the pipeline.

GPU workloads

Clusters for model training and inference, node pools, scheduling and model rollout pipelines.

What you keep

  • Reference architecture and cluster audit
  • Clusters and add-ons as code
  • GitOps repository structure
  • Upgrade runbook

Stack

  • Kubernetes
  • Helm
  • ArgoCD
  • Terraform
  • Vault
  • Prometheus
  • GPU
git log --cases

Similar problems we have solved

All case studies
Fintech and EDI

SberCorus

Releases were rare and manual; every deployment was an event.

  • Moved services to Kubernetes with zero-downtime updates
  • Built build, test and deploy pipelines for every environment
  • Cleaned up secrets management and access rights
AI platforms

JedAI

ML workloads need different infrastructure than regular web services.

  • Deployed clusters for training and inference
  • Made environments reproducible
  • Built model rollout pipelines
System integration

IMSystems

Deployments for different customers in air-gapped environments had to be repeated by hand.

  • Described the infrastructure as code so every environment deploys the same way
  • Accounted for air-gapped networks with limited internet access
  • Cut the time needed to stand up a new environment

TEAM.md

Who will work on your project

There are three of us, by design: the engineers you talk to are the engineers who do the work. No account managers, no hand-offs.

  • Alexander Kaloshin — Co-founder

    Alexander Kaloshin

    Co-founder

    20+ years in IT. In 2014 he founded Last.Backend, a container orchestrator built at almost the same time as Kubernetes. Leads our AI practice.

    • Last.Backend: container management and DevOps as a service
    • Fault-tolerant infrastructure: Kubernetes, CI/CD, bare metal
    • Go engineering
    • Speaker at ProIT Fest and GolangConf
  • Alexander Tikhomirov — Co-founder

    Alexander Tikhomirov

    Co-founder

    7 years in DevOps and dozens of delivered projects, from a first pipeline to production under heavy load.

    • Infrastructure design and operations
    • Cloud migrations and greenfield builds
    • CI/CD, Kubernetes, infrastructure as code
  • Denis Gerasimov — Co-founder

    Denis Gerasimov

    Co-founder

    5 years in DevOps. Takes on the hardest problems and sees them through.

    • Complex infrastructure projects
    • Monitoring, on-call and incident response
    • Operations automation
FAQ.md

FAQ

Will you tell us if we don't need Kubernetes?
Yes. For a handful of services, a managed container platform or plain VMs can be cheaper and simpler. We'll say so in the audit.
Managed Kubernetes or self-hosted?
Managed by default in public cloud. Self-hosted when you're on-premises, air-gapped or have strict data-residency rules — we've done both.
Can you run Kubernetes in an air-gapped environment?
Yes. We build offline registries, mirror charts and images, and set up update procedures that don't need internet access.
Do you only consult, or also operate?
Both. Many clients keep us on a retainer with 24/7 support after the build.
Do we own everything you build?
Yes. Code, Terraform, Helm charts and runbooks live in your repositories from day one, and cloud accounts stay in your name. If you ever take the work in-house, the handover is documentation and a call, not a rescue project.

Let's talk about your infrastructure

Tell us what is going on today. We reply within one business day with a free initial review.

Talk to an engineerServices