At Publicis Re:Sources, I architect and maintain ETL/ELT pipelines with Azure Data Factory and Azure Databricks, bringing data from REST APIs, SQL databases, and blob storage into a Delta Lake Lakehouse.
I optimized PySpark scripts, Delta table partitioning, and cluster resources, reducing pipeline query latency by 35% and cloud compute overhead by 20%. I also built Python frameworks for automated API ingestion.
I engineer multi-agent workflows with LangGraph and LangChain, and designed RAG systems using Pinecone that reduced internal document search response times from minutes to sub-second. My work also includes machine learning classifiers and regressors for marketing campaigns and customer behavior forecasting.
