Description
ed design and delivery of a scalable predictive risk classification platform to detect counterfeit, expiry, and invalid-order risks across the supply chain. Integrated SAP S/4HANA, ICH and ATTP sources with Azure pipelines (ADF, Databricks, ADLS), Azure ML scoring, and SAP BTP API actions. Built OLAP and real-time insights (Synapse, Power BI) with end-to-end observability and security, driving ~$3M annual savings and measurable reductions in recalls and manual triage.
Roles and Responsibilities
Owned data engineering architecture and end-to-end solution design for the SCM alerts platform. Designed and implemented ADF ingestion pipelines to process 10M+ alerts/day and orchestrated Databricks (PySpark) enrichment. Defined data lake schema and partitioned Parquet strategy in ADLS for scalable storage and query performance. Developed Azure ML scoring workflows and productionized models; integrated inference results into SAP via BTP APIs to trigger automated actions. Built Synapse-based OLAP/real-time layers and Power BI dashboards with role-based access for operational and executive insights. Implemented observability, monitoring, and security controls (Azure Monitor, Key Vault, logging/telemetry). Tuned pipeline performance, managed costs, and enforced data governance and compliance for post-demerger separation. Coordinated cross-functional stakeholders, led deployments, and validated business impact (20% fewer recalls, 30% fewer invalid orders, 50% reduction in manual triage).
Technologies: SAP S/4HANA, ICH, ATTP, SAP BTP (APIs), Azure Data Factory (ADF), Azure Databricks, Apache Spark, PySpark, Python, Azure Data Lake Storage (ADLS), Parquet, Azure Machine Learning, Synapse Analytics, Power BI, Azure Monitor, Azure Key Vault
Industry: Pharmaceutical