Muhammad Osama

Led client-facing solution architecture as a Part-time Solutions Architect, directing 5 DataOps engineers. Aligned technical execution with business priorities via weekly syncs with operations and stakeholders. Led upgrade and re-architecture of analytical Vertica clusters (~800 TB data, 400+ entities) to improve performance, stability, and auditability, achieving 20-30% faster queries and downtime under 2 hours per maintenance. Developed fallback, backup, restore, and validation protocols ensuring zero data loss in multi-node failure scenarios. Bridged engineering and client delivery teams, owning each solution phase from design through rollout.

Architected end-to-end Medallion-architecture data platform using dbt + Snowflake; implemented Data Vault 2.0 (Silver) and Kimball (Gold) across Transactions, Orders, Payments, Users. Embedded GDPR governance with PII masking and encryption at the Snowflake layer, enabling MaRisk and AML compliance. Orchestrated dbt DAGs via Airflow with 70+ data quality tests and 95.5% SLA via automated dashboards. Implemented ephemeral Snowflake schemas per PR via GitHub Actions for testing and drift detection. Collaborated with Product/BI to validate semantics and guide schema optimization, retiring legacy MS SQL Server infrastructure.

AdTech platform optimization with ontology-driven modeling. Refactored Spark pipelines reducing processing costs by 30% (~€70K/year) and runtime by 45%. Unified multi-source campaign data (XANDR, GAM, Adobe Analytics) enabling self-service analytics and ~30% ad ROI improvement. Designed ontology-driven data models improving cross-team discovery by 60% and reducing duplicates by 40%. Engineered self-optimizing orchestration with adaptive provisioning achieving ~20% better cluster utilization.