Data platform Architect

Data Platform Architect

Location: Brussels, hybrid

Level: Senior / Principal

Languages: English required; French a plus

About the Team

We're building a data mesh on Azure and Databricks: domain teams own their data as products, and the central platform team gives them a self-serve, governed environment to do it well. We have an MVP platform where pilot teams are already onboarding. The next phase is about evolving it: hardening what works, reworking what doesn't, and building the capabilities the next wave of domains will need.

The Role

You'll take ownership of the platform, its architecture and its evolution roadmap, and lead the team that builds it. That starts with an objective assessment of the current setup — what to keep, what to strengthen, what to rework — and turns into a roadmap, shaped with the team and stakeholders, that balances stabilisation, cost and new capabilities.

Databricks is at the core of the platform, so deep hands-on expertise there — Unity Catalog, workspace topology, Delta, streaming, cost and performance — is essential rather than one skill among many. It's a senior role that mixes architecture, technical leadership and hands-on work — expect a significant amount of your time in the codebase, spiking hard problems and reviewing what the team ships.

We value demonstrated experience to lead the design and implementation over familiarity with the concepts.

What You'll Do

• Assess the current platform objectively — architecture, automation, cost, operations — and turn that into a prioritised roadmap: what to stabilise, what to rework, and what to build next.
• Own and evolve the end-to-end platform architecture — ingestion, storage, transformation, streaming, serving and governance — designing from explicit ASRs and treating cost, resiliency, automation and security as design inputs rather than afterthoughts.
• Own and evolve the Databricks architecture: Unity Catalog design, workspace and catalog topology per domain, compute and cluster policy strategy, Delta and streaming patterns, and the performance and cost standards the team builds to.
• Evolve the platform incrementally: strengthen or rework existing components without disrupting the domains already running on them, and deliver new capabilities in a steady cadence.
• Evolve the data product landing zone: a repeatable, isolated environment per domain covering resource organisation, networking, identity, access boundaries and Databricks workspaces — and guide the team in automating its delivery end to end, so onboarding a new domain becomes routine and new platform features ship without manual builds.
• Define the platform security architecture with the security team: identity and access model, network isolation and private connectivity, secrets, classification and least privilege — designed to pass audit without slowing teams down.
• Drive the data mesh operating model: domain ownership, data product contracts, federated governance, and the interoperability standards and shared definitions that let autonomous domains combine their data reliably.
• Architect FinOps, observability and data quality as platform capabilities:
• FinOps — every domain can see what its data products cost and act on it.
• Observability — mesh health is visible in aggregate, and the impact of a degraded data product is known before consumers report it.
• Data quality — domains own their checks, but the platform provides the mechanism: a sidecar-style capability with scheduled execution, results collected centrally and surfaced in the data governance tool alongside lineage and ownership.
• Ensure the platform integrates with the enterprise data governance tooling [Collibra / Unity Catalog] so catalogue, lineage, quality and ownership are visible in one place.
• Set and hold infrastructure-as-code and automation standards in Terraform through review — you define the patterns; the team builds most of it.
• Produce architecture documentation that makes complex design clear to engineers, reviewers and business stakeholders, and take it through architecture review.
• Lead the platform team: technical direction, backlog prioritisation, thorough design and code reviews, and hands-on support when engineers are blocked.
• Work with data product teams and business stakeholders to understand their needs and shape a sequenced platform roadmap.

What Success Looks Like in the First Year

• Ownership established — you are recognised by the team and stakeholders as the owner of the platform's architecture and direction, and decisions route through you.
• Assessment and roadmap — a clear, evidence-based view of the current platform and an agreed evolution roadmap, shared with and backed by the team and domain stakeholders, within the first quarter.
• Smooth onboarding — new domains land on an automated, governed environment in days rather than weeks, with no bespoke work per domain.
• Stabilisation — the existing platform runs predictably: fewer incidents, known failure modes handled by automation, and clear operational ownership.
• Cost optimisation — measurable reduction in platform and per-data-product spend, with cost visible to every domain.
• New capabilities delivered — the priority features on the roadmap [e.g. streaming, observability, quality] shipped and adopted by domain teams.

What We're Looking For

• 8+ years in data engineering or platform roles, including 3+ years architecting production data platforms — specific systems, specific decisions, specific outcomes.
• Someone who can own the design, implementation and roadmap of a platform through a shared vision — assessing what exists, deciding what to keep and what to rework, aligning the team and stakeholders behind the direction, and delivering change without disrupting live users.
• Strong Azure across the data stack and its foundations: ADLS, Event Hubs, networking (VNets, private endpoints, DNS), identity and RBAC, Key Vault, resource organisation and policy. AWS a bonus.
• Experience designing landing zones or isolated multi-team environments, including Databricks workspace topology, and automating their provisioning.
• Deep, hands-on Databricks expertise — essential. Unity Catalog design and governance, workspace and catalog topology for multi-team setups, Delta Lake, Workflows, Structured Streaming, cluster policies and compute strategy, system tables, performance and cost tuning. You should be able to walk through Databricks platforms you've designed and operated.
• Experience designing security and governance in a cloud data platform that has survived audit.
• Experience building self-serve, domain-oriented data platforms; hands-on data mesh implementation strongly preferred.
• Delivered cost visibility, observability or data quality capabilities on a platform, with results you can point to; experience integrating with a data governance/catalogue tool.
• Near real-time architecture and delivery with Event Hubs, Kafka and Spark Structured Streaming, with a firm grasp of the trade-offs against batch.
• Strong Terraform and automation discipline.
• Clear architecture documentation and the ability to defend a design under challenge.
• Technical leadership experience: leading a team, prioritising work, mentoring and reviewing.
• Comfortable in the codebase — Python, SQL, Spark, Terraform.

Nice to have: AWS, regulated-industry experience.

What We Offer

• A platform-as-a-product culture where adoption by domain teams is the measure of success.
• Real autonomy over architecture decisions, and a team that values working code and clear documentation over slide decks.