Aug 2020 – May 2023 · 2 yrs 9 mos · New York, United States · On-site

Oct 2017 – Jul 2020 · 2 yrs 9 mos · Washington, United States · On-site

Dec 2015 – Sep 2017 · 1 yr 9 mos · Atlanta, Georgia, United States · On-site

• At UPS, I designed and developed real-time and batch ingestion pipelines on AWS, capturing millions of daily package scans, IoT telematics, and GPS events. Real-time pipelines used Kinesis Data Streams, Firehose, and Lambda, landing data into S3, while batch pipelines leveraged AWS Glue (PySpark), Step Functions, and Lambda to consolidate transportation, warehouse, and route planning data into a centralized S3 data lake with schema evolution handling.
• I modeled curated datasets in Redshift and Aurora PostgreSQL using star and snowflake schemas, supporting reporting on transportation costs, service-level KPIs, and fleet utilization. Spark ETL pipelines on EMR with Hive applied transformations, custom UDFs in Scala/Python, and enrichment rules to standardize and cleanse delivery event streams. Partitioning, bucketing, ZSTD compression, and Parquet/ORC formats optimized storage and query performance.
• Workflows were orchestrated using MWAA (Managed Airflow) and Step Functions, with DAG dependencies, backfilling, and failure recovery via SNS alerts. Data quality frameworks with PyDeequ, Great Expectations, and Python scripts ensured completeness, timeliness, and anomaly checks. Feature pipelines were delivered into SageMaker Feature Store for ML models predicting delivery ETAs, demand, and route optimization.
• I built dashboards in Kinesis Analytics and QuickSight, providing visibility into deliveries, warehouse bottlenecks, and exceptions. CI/CD pipelines were automated using CodePipeline, CodeBuild, Terraform, and GitHub Actions. Governance was enforced with Lake Formation, IAM roles, Redshift security, and S3 encryption (GDPR/CCPA). EMR clusters were tuned for performance and cost using auto-scaling, spot instances, and Spark optimizations. These efforts enabled scalable, automated, and secure cloud-native data pipelines that enhanced operational efficiency, analytics reliability, and real-time visibility across UPS logistics operations.

May 2014 – Oct 2015 · 1 yr 5 mos · India · On-site

• At UPS, I designed and optimized large-scale data engineering solutions on AWS to support logistics and supply chain analytics. I analyzed ETL pipelines, reviewed SQL queries, and tuned Amazon Redshift workload management queues to remove inefficiencies, achieving up to 40% faster runtimes. By strategically applying distribution/sort keys, compression encoding, and partition pruning, I improved scalability and reduced latency for millions of daily package events.
• I built and supported enterprise-grade data warehouses on Redshift and Snowflake, designing dimensional models with star/snowflake schemas and implementing Slowly Changing Dimensions (SCD2) for historical accuracy. Using Amazon S3, AWS Glue (PySpark), and Redshift Spectrum, I developed automated pipelines integrating structured and semi-structured data. Orchestration with AWS Step Functions and Lambda reduced manual effort, while PyDeequ, Great Expectations, and CloudWatch monitoring with SNS/Slack alerts ensured data quality and SLA compliance.
• I consolidated data ingestion across transactional systems, APIs, and legacy databases with AWS Glue, Informatica, and Data Pipeline, standardizing processes for reliability. Security and governance were enforced using IAM roles, RBAC, and KMS encryption, ensuring compliance with GDPR/CCPA.
• I also optimized query performance in Redshift and Snowflake through indexing, clustering, and compression, enabling interactive analytics in Amazon QuickSight dashboards with automated refreshes. Automated Informatica sessions and Glue workflows with Lambda and Step Functions improved resilience, scalability, and monitoring with CloudWatch/X-Ray logging.
• These efforts delivered a secure, high-performance, and automated cloud-native data ecosystem that streamlined ingestion, accelerated analytics, reduced costs, and improved operational visibility across UPS’s logistics network.