Mohammed Sr.

Modernized the Federal Subaward Reporting System (FSRS) by migrating high-volume workloads from legacy MySQL to cloud-native distributed frameworks. Engineered ETL/ELT pipelines and CDC mechanisms spanning AWS, GCP, Snowflake, and Databricks to support near–real-time synchronization of millions of federal records. Built Spark and Glue-based ingestion and transformation workflows, designed Lakehouse patterns with Iceberg/Delta, and delivered dbt modular transformations with testing and lineage. Implemented AWS MWAA Airflow DAGs, OpenSearch indexing, and Terraform-based infrastructure as code to ensure consistent, secure, audit-ready environments. Improved performance through SQL and Spark tuning, implemented cross-environment validation checks, and supported operational troubleshooting for data discrepancies. Also supported re-platforming from Redshift to Snowflake and prototyped transformations in Databricks for production readiness.

Built and evolved data pipelines that transform policy and claims data into analytics-ready datasets on AWS. Developed ML feature layers in collaboration with data science teams for fraudulent-claims detection, including end-to-end testing and environment validation. Implemented ETL using Python/PySpark, Matillion-to-Snowflake loading, and orchestration using AWS services such as Step Functions for SageMaker workflows. Designed lakehouse patterns and medallion architecture in Databricks with Delta Lake (including SCD handling), integrated MLflow for experiment tracking, and authored dbt models and tests on Snowflake to standardize claim KPIs. Migrated legacy scheduling approaches to Airflow DAGs and used Terraform to provision environments and data platform components with reliability and governance in mind.

Modernized legacy Oracle systems to AWS and delivered batch and streaming ETL pipelines for vaccine data ingestion and reconciliation. Built PySpark transformations on AWS Glue to cleanse and validate immunization records, designed Redshift star schemas, and optimized SQL for reconciliation reporting and duplicate detection. Orchestrated pipelines using Step Functions, Lambda, and messaging services with robust error handling and retries. Enabled near real-time ingestion through Kafka and supported CDC-compliant reporting with partitioned incremental loads. Added data quality automation using Great Expectations, built secure inter-agency APIs, and monitored production reliability with CloudWatch and Grafana dashboards. Supported CI/CD with Jenkins and GitLab CI and used Terraform to codify infrastructure across dev/QA/prod environments.