Qbitech is looking for a Senior Data Architect to take technical ownership of a large-scale data platform migration for an international consumer technology organisation operating in a heavily regulated market. This is a modernisation programme, not a greenfield build.

The scale is real. The current platform is an on-premise Oracle data warehouse where the core fact table alone holds tens of billions of rows and hundreds of gigabytes, on top of multi-year history carrying strict retention obligations in several domains. It is fed by more than 150 legacy integration points — hundreds of Oracle database links plus dozens of API and file feeds — with transformations running as PL/SQL batch and hourly jobs. Around 170 tables are being moved onto change data capture streaming through Kafka. Source systems span the core transactional platform, financial ledger, clickstream, marketing automation, mobile attribution, customer experience data and a range of third-party providers.

The high-level target architecture is already defined, and you would be building within it rather than choosing it. The direction is set: an AWS lakehouse using Apache Iceberg for storage, dbt for transformations, Apache Airflow for orchestration, Trino for federated SQL, ClickHouse for near-real-time analytics, Redshift for traditional BI workloads, and a semantic layer and data catalog on top, following a medallion pattern across bronze, silver and gold layers. Layer responsibilities and delivery flow are documented. What is genuinely still open — and yours to shape — is the data contracts and SLA framework between teams, the long-term event-driven ingestion architecture that replaces today's tactical database links, how governance standards get enforced, cost optimisation and performance tuning, and the Redshift integration strategy.

This is a hands-on role as well as a design role, and the split is roughly 60% hands-on to 40% strategic. You would write dbt models as reference implementations and proof-of-concepts, work directly in AWS across S3, Athena and Lake Formation, and sit close to the engineers building the platform. The other 40% is genuine architecture ownership: setting technical standards, running architecture workshops and design reviews as the design validator, authoring architecture decision records, and mentoring the team.

You would sit inside the Data Platform Engineering team as the technical authority for data architecture, owning design patterns, ADRs and technology strategy for the data domain independently, and collaborating with the Enterprise Architecture team on cross-cutting concerns. A core delivery team of around 12 engineers builds against your designs, with 30 to 40 more across analytics, data science and data engineering contributing to the programme.

Machine learning is already running in production across several use cases, so part of the platform charter is making it AI-ready — feature stores, real-time data access and MLOps integration patterns — rather than building models yourself.

The programme runs to a firm multi-year deadline with executive sponsorship, so this is business-critical work with real momentum behind it.