Mamatha Velturi

About Mamatha Velturi

With 10+ years of experience in Data Engineering and ETL development, I specialize in building and optimizing scalable data solutions using GCP, Azure, and AWS. I have a strong track record in designing robust data pipelines, implementing modern data architectures (Delta Lake, Data Mesh), and ensuring efficient data migration and transformation processes. I am proficient in implementing robust ETL processes, leveraging tools such as Big Query Data Transfer Services (DTS), Google Cloud Dataflow, IICS, SSIS, Talend, and Databricks Pipelines, and have extensive experience with data modeling, data warehousing, and managing large-scale data solutions in the financial and banking sectors. Additionally, I have a strong background in working with open table formats like Iceberg and Data Lake, ensuring optimal data storage and retrieval.

Experience

1. Sr Etl Developer

   Capital One

   Oct 2024 — Present

   Spearheaded a cloud migration project, moving critical data from on-premises SQL databases to Cloud, which resulted in a 50% reduction in storage costs and improved scalability for processing trial data.Developed automated data validation and error-handling mechanisms to ensure compliance with HIPAA regulations, improving overall data accuracy.Designed and developed ETL/ELT workflows using AWS Glue, Step Functions, and Lambda to automate ingestion and transformation of auto loan and lease data.Amazon Kinesis Data Streams were utilized to capture real-time loan application status updates from dealer portals.Designed CloudWatch dashboards, custom metrics, and alarms for proactive monitoring of ETL jobs and AWS resource utilization.Provided support to the Auto Finance division by developing scalable solutions for loan origination, refinancing, and lease servicing utilizing AWS.Conducted regular ETL performance tuning, incorporating indexing, partitioning, and parallelization techniques to ensure timely data processing and reduce processing times in the data pipeline by 35%.Created, trained, and deployed end-to-end ML models (classification, regression, NLP, and forecasting) with Python, scikit-learn, PyTorch, and TensorFlow on AWS infrastructure.Implemented end-to-end ML workflows in AWS SageMaker, including model training, tuning, and production hosting.Implemented feature engineering pipelines with AWS Glue and PySpark to condition model input data for performance and interpretability.Created custom algorithms in SageMaker with containerized frameworks using Docker and ECR (Elastic Container Registry).Automated retraining and testing of models using SageMaker Pipelines in collaboration with Step Functions and EventBridge.

Education

• Jawaharlal Nehru Technological University

  Master of Technology - MTech, Engineering Technology, General