[01] · SERVICES 4 CORE AREAS + 2 FORMATS

What we do.
And what we don't.

We do not take engagements outside the four core areas below. With a team of 14 we have to pick our battles — and that is on purpose.

01 / 04 · CLOUD ARCHITECTURE

Cloud architecture

We set up cloud landscapes so nobody has to improvise later. AWS and GCP focus, Azure on request. Greenfield, heavy migration, or remediation of grown structures.

Landing zones & multi-account architectures (AWS Control Tower, GCP Org)
Networking, IAM, security baselines (CIS benchmarks, SCPs, IAM boundaries)
Cost engineering & FinOps fundamentals (tagging, showback)
Migration strategy (6R assessment) including cutover support
Disaster-recovery and backup designs with drills

Preferred stack: Terraform · AWS Organizations / GCP Org Policies · IAM Identity Center · Transit Gateway / Cloud Interconnect · Cloud Trail / Cloud Audit · Vault.

02 / 04 · PLATFORM ENGINEERING

Platform engineering

Internal developer platforms make the difference between „we ship features" and „we wrestle YAML". We build the layer your engineering teams need — self-service, security, observability, all in production-ready form.

Kubernetes platforms (EKS, GKE, on-prem) with operator pattern
GitOps workflows (ArgoCD / Flux) including multi-cluster promotion
Internal developer portals (Backstage, Port) tailored to your actual workflow
Observability stack: Prometheus / Mimir / Loki / Tempo / Grafana
Service catalog, templates, scaffolding

Preferred stack: Kubernetes · ArgoCD · Backstage · Crossplane · Cert-Manager · OPA / Kyverno · Prometheus + Grafana · OTel.

03 / 04 · DEVELOPER EXPERIENCE

Developer experience

Velocity does not come from new tools — it comes from measurably less friction in the daily workflow. We measure first (DORA, build times, MTTR), then optimise.

CI/CD audits and refactoring (GitHub Actions, GitLab CI, BuildKite)
Local development environments (devcontainers, Tilt, Skaffold)
Build performance (Bazel, Buck2, BuildKit, caching layers)
Test strategy (pyramid → diamond), flake reduction
Monorepo setups (Nx, Turborepo, pnpm workspaces)

Preferred stack: GitHub Actions · BuildKit · Nx / Turborepo · devcontainers · Bazel. Measurement: DORA + SPACE framework.

04 / 04 · DATA INFRASTRUCTURE

Data infrastructure

Data platforms designed to last five years — not 18 months. We design ELT pipelines, lakehouses and modern-data-stack setups so analytics and operational workloads stay cleanly separated.

ELT pipelines with dbt + Airflow / Dagster
Warehouse / lakehouse: Snowflake, BigQuery, Databricks
Streaming: Kafka, Flink, ClickHouse for real-time use cases
Data contracts & quality checks (Great Expectations, dbt tests)
Knowledge transfer to your in-house data team

Preferred stack: dbt · Snowflake / BigQuery · Airflow / Dagster · Kafka · ClickHouse · Iceberg · Great Expectations.

[+] / FORMATS

Two delivery formats across the four disciplines.

FORMAT A4 WEEKS · FIXED

Engineering audit

A senior engineer and an architect spend 4 weeks inside your organisation. Output: written assessment + prioritised roadmap with effort estimates — 80–120 pages of substance, not a 12-slide summary. Fixed price.

Weeks

Engineers

FIX

Pricing

FORMAT B3–6 MONTHS

Embedded team

2–4 of our engineers join your squads. They write code, do reviews, take on-call shifts — and hand over cleanly at the end. Time-and-material, monthly cancellation after a 3-month minimum.

3–6

Months

2–4