Sonia Hayes

I’m a Principal Data Engineer and Data Architect with 10+ years designing enterprise-grade, multi-cloud lakehouse and distributed data platforms across healthcare, fintech, and enterprise domains. I focus on exactly-once streaming, CDC frameworks, metadata-driven ingestion, Data Vault modeling, and governance standards including HIPAA, HITRUST, and SOC2.

I’ve built AI-enabled data foundations using feature stores, embedding pipelines, vector search, drift detection, and RAG-ready architectures across Databricks, Snowflake, Kafka, and Spark ecosystems. I also lead observability-driven DataOps, Spark performance tuning, partitioning strategies, and cost optimization—delivering 99.99% platform availability at scale.

Recently at DataForest, I architected multi-cloud lakehouse ingesting 6TB+ daily across 140+ systems, designed exactly-once Kafka pipelines sustaining 75K events/sec peak throughput, and reduced Spark shuffle skew 39%. Earlier, at CorroHealth, I built HIPAA/HITRUST/HITECH-aligned FHIR/HL7 lakehouse processing for 2.7B+ claims and clinical records monthly, strengthening security with PHI masking, encryption-at-rest, and audit logging while reducing warehouse compute costs 28%.