Kundan Kumar

Architected a HIPAA-compliant AWS lakehouse on S3 using Apache Hudi to process genomics and clinical data with a <10-minute SLA, using CDC-based ingestion with idempotent writes and schema evolution. Reduced production pipeline failures by ~60% through checkpointing, schema validation, and automated failure recovery, and optimized PySpark workloads to improve throughput.

Designed hybrid batch and streaming pipelines for clinical analytics (~1 TB/day), reducing ETL latency by 50% via Spark execution-plan optimization and partition redesign. Built FHIR/HL7 real-time ingestion into a unified clinical data store, improving observability and reducing MTTD by ~40%, while implementing data quality frameworks for AI/LLM workflows.

Engineered distributed ETL and streaming pipelines processing 30–40 TB/day of semiconductor telemetry, XML, and image data across global manufacturing systems. Reduced batch processing time by ~50% using adaptive query execution and join tuning in Apache Spark, and built Kafka-based streaming for sub-minute anomaly detection, reducing RCA time by ~40%.

Built a patient-practitioner engagement platform supporting report management, slot booking, payments, prescriptions, and discharge summaries for 50,000+ patients. Implemented HIPAA-compliant FHIR/HL7 interoperability into a centralized repository and developed RESTful APIs and data transformation pipelines, integrating clinical AI safety checks to reduce manual medication review time by ~30%.

Grade: 8.9 / 10

Earned an M.Tech in Computer Science from NIT Calicut, graduating with a GPA of 8.9/10.

Grade: 8.5 / 10

Activities and societies: Gold Medallist; University Rank 1

Completed a B.Tech in Information Technology at Maharshi Dayanand University, graduating with a GPA of 8.5/10 and receiving University Rank 1.