Cloud native · Kubernetes · AI

Cloud native, Kubernetes and AI.Built for production.

OKKO is a Belgian IT consultancy founded in 2018. We design, deploy and evolve your cloud platforms and AI projects, commercial or sovereign, alongside your teams rather than instead of them, from Liège to Brussels and across the Benelux.

We work with

  • Kubernetes
  • AWS
  • Microsoft Azure
  • Google Cloud
  • European clouds
  • On-premise

Team certifications

  • CKA
  • CKS
  • AWS Solutions Architect Professional

Services

Three areas of expertise, one standard: it has to run.

We work where infrastructure, software and AI meet, from scoping to production, and for the long run.

  • Cloud native & Kubernetes

    Design, migrate and operate Kubernetes platforms that are reliable, secure and cost-efficient, on AWS, Azure, Google Cloud, a European cloud or your own servers.

    • Architecture & migration
    • Platform engineering & GitOps
    • Observability, security, FinOps
  • Sovereign & commercial AI

    Put generative AI to work for your business, with market-leading models or open-weight models hosted on your premises or in Europe. Your data stays under your control.

    • Open-weight LLMs on-premise or in the EU
    • RAG & internal assistants
    • Agents & automation
  • Modernisation & DevOps

    Help your development teams ship faster and with more confidence on the cloud, and bring AI into your products without rewriting what already works.

    • Containerisation & modernisation
    • CI/CD & DevOps practices
    • AI integration in your applications

Approach

Analyse, deliver, iterate.

No six-month tunnel, no report gathering dust. We move in short, measurable increments, and every step leaves something usable in production.

  1. 01

    Analyse

    We start from your reality: architecture, teams, regulatory constraints and budget. A short assessment, clear priorities.

  2. 02

    Deliver

    We build with your teams, in small batches. Everything is versioned, documented and automated: Infrastructure as Code, GitOps, CI/CD.

  3. 03

    Iterate

    We measure, adjust and hand over. Your needs change; the platform and the organisation follow, without starting from scratch.

  • Open source and open standards first
  • No artificial lock-in to OKKO
  • Transparency on choices and costs

Working together

Four ways to get started

Depending on your maturity and your teams, we take on a short, focused scope or stay for the long haul.

  • Audit & assessment

    A few days to a few weeks to assess a platform, a cluster or the feasibility of an AI project.

    You getA clear picture and a prioritised, costed roadmap.

  • Fixed-scope project

    One scope, one deliverable, one commitment: a migration, a platform, an AI use case from prototype to production.

    You getA solution in production, documented and handed over.

  • Team augmentation

    Senior engineers embedded in your teams to speed up a project or bring in expertise you are missing.

    You getImmediate capacity, and teams that level up.

  • Managed operations

    We operate and evolve your Kubernetes and AI platforms, with service levels we define together.

    You getMonitoring, upgrades and support at the agreed service level.

Case studies

What we have delivered

A few recent engagements. Names stay confidential; results do not.

  • Video broadcasting

    Embedded Kubernetes, fully offline

    Context
    No prior experience with container orchestration, and an ambitious goal: Kubernetes clusters embedded in trucks that operate anywhere in the world, from jungle to desert, with no connectivity.
    What we did
    A tailor-made, fully air-gapped Kubernetes cluster, with performance tuning for video workloads: transcoding, labelling and more.
    Outcome
    A self-contained Kubernetes platform that runs without any connection, for a client starting from scratch.
  • Financial institution

    Secure cloud landing zone

    Context
    No landing zone or least-privilege model yet, and cloud authentication to integrate with the corporate identity system.
    What we did
    Landing zone set-up, integrated SSO, and policies (guardrails) that enforce internal rules automatically.
    Outcome
    More than 800 applications now run on this landing zone, which keeps evolving to add new services securely and compliantly.
  • Public administration

    Sovereign AI assistant for staff

    Context
    Staff needed to analyse documents and draft answers from internal knowledge bases, without data leaving the organisation.
    What we did
    Local AI on on-premise hardware, RAG over internal databases and knowledge bases, MCP servers to query in-house software.
    Outcome
    A six-month project overall, with the MVP delivered at two thirds of the estimated budget.

Technology

Proven tools, chosen for you.

We favour open source and portable architectures, so you stay free to choose.

Platforms
  • Kubernetes
  • OpenShift
  • Helm
  • Argo CD
  • Flux
  • Cilium
Cloud & IaC
  • AWS
  • Azure
  • Google Cloud
  • OVHcloud
  • Scaleway
  • Terraform
  • OpenTofu
Observability & security
  • Prometheus
  • Grafana
  • OpenTelemetry
  • Loki
  • Vault
  • Keycloak
AI
  • vLLM
  • Ollama
  • Mistral
  • Llama
  • Qwen
  • LangGraph
  • pgvector
  • Qdrant

Careers

You like it when things run in production?

We are looking for curious, demanding engineers (platform, cloud, software, AI) who would rather solve real problems than fill slides.

  • Varied engagements, from scale-ups to the public sector
  • Senior peers and time to learn
  • A small team where your opinion matters

FAQ

Frequently asked questions

What do you mean by “sovereign AI”?

AI where you control the models, the infrastructure and the data. In practice: open-weight models (Mistral, Llama, Qwen…) hosted on your servers or with a European provider, with no data sent to a third party. It is often the prerequisite for using AI on sensitive, regulated or confidential data.

Do we have to choose between commercial and sovereign AI?

No. You can combine both: commercial APIs for low-sensitivity use cases, hosted models for critical data. We help you decide case by case, based on quality, cost, latency and compliance (GDPR, AI Act).

Do you need Kubernetes for an AI project?

Not always. For a prototype, a managed service is often enough. Once you need to serve models at scale, share GPUs or industrialise, Kubernetes becomes a real asset, and that is our home turf.

Which clouds do you work with?

AWS, Microsoft Azure and Google Cloud, European clouds such as OVHcloud or Scaleway, and on-premise. We favour portable architectures to avoid vendor lock-in.

How long does a first engagement take?

An audit usually takes a few days to a few weeks. It ends with a costed roadmap; you then decide what comes next, with us or in-house.

Do you work with other service companies?

Yes. We work as reinforcement or subcontractor for IT service firms and integrators that need deep cloud native or AI expertise.

Contact

Let's talk about your project

Tell us about your context in a few lines. We will get back to you within two business days with a first take and, if it makes sense, a meeting proposal.

Prefer a direct line?

contact@okko.be