Senior Data Architect (Azure/Databricks/Cosmos DB)

Interview process: Hr call 30min, 1 intro call + 1 technical interview (1 hour)
Location EU
Timezone: CET
English: C1 ideally

Critical Note
This role is for an Architect, not a Data Engineer. Candidates who have only implemented pipelines without owning end-to-end architecture decisions will not be considered — this is explicitly called out multiple times by the client.

Project Overview
Client needs a Data Architect to take ownership of an existing, complex, partially problematic Azure data platform — assess it quickly, separate tactical fixes from long-term architecture, and produce implementation-ready recommendations that engineering teams can act on immediately.

What You'll Do
Design, build, and optimize scalable data architectures on Databricks + Azure
Own end-to-end architecture decisions — not just implement pipelines
Assess an existing, incompletely documented platform and deliver concrete, actionable output within the first 1–2 weeks
Design scalable, reusable, metadata/configuration-driven ingestion frameworks (multiple databases, APIs, file-based sources)
Work hands-on with Azure Cosmos DB — data modeling, partition-key strategy, indexing, RU/throughput optimization, scaling, consistency, Change Feed
Work across Synapse workload patterns: Pipelines, Spark, Serverless SQL, Dedicated SQL Pools
Handle genuinely unstructured data — documents, PDFs, scans, free text, images, emails (real examples required, not theoretical)
Make technology-neutral architecture calls — comfortable concluding Databricks/Synapse/ADF may NOT be the right fit if scale/cost/requirements don't justify it
Separate and progress tactical stabilization vs. long-term target architecture in parallel
Produce clear architecture documentation, diagrams, and technical recommendations for engineering teams
Collaborate with engineering, analytics, and business stakeholders; mentor junior team members

Must-Have Requirements
Architecture ownership — proven, concrete examples of personally designing/owning a data platform architecture (not just pipeline implementation)
Azure — hands-on, recent (not years-old), 4+ years minimum. Azure DevOps usage alone does NOT count as Azure cloud experience
Databricks — must have real hands-on experience
Azure Cosmos DB — must have hands-on experience (scaling, partitioning, RU optimization, advanced design decisions) — this is one of the most heavily emphasized requirements in the entire brief
Azure Synapse — depth across multiple workload patterns, not surface-level exposure
Unstructured data — proven, real examples (not just structured/relational data experience)
Metadata/configuration-driven ingestion — must be able to explain a reusable ingestion framework design in depth
Existing platform modernization — concrete example of inheriting a problematic platform and improving it
Python, SQL, ETL/ELT
Strong data modeling (fact tables, dimensional/star schemas)
Batch + incremental processing/refresh
Strong communication with both technical and business stakeholders
CV must contain clear, detailed hands-on Azure experience — vague CVs will result in automatic rejection

Nice to Have
Future-oriented thinking on AI / natural-language querying of enterprise data
Experience with legislation/regulated data platforms (client referenced a similar past project: legislation data platform on Azure Synapse — XML/API pipelines, medallion architecture)

Interview Prep Note (for candidate)
Client wants 2–3 strong real project examples structured as: Problem → options considered → personal decision → why → implementation → result. Expect deep technical probing on Cosmos DB and Synapse specifically, plus a scenario where Databricks is deliberately removed from the discussion to test technology-neutral thinking.