Professional Experience
Data Scientist
Fannie Mae - Washington, D.C | August 2024 - March 2025
Contract through: Systems Engineering Services Corporation
Team : Enterprise Modeling and Analytics
- Migrated a Single-Family mortgage data pipeline from Python to PySpark, reducing execution by 40%, improving scalability, and enabling smooth production deployment in a distributed AWS environment.
- Designed and implemented a cost-saving Amazon EFS storage lifecycle model to automate transitions between storage tiers (Standard, Infrequent Access, and Archive), resulting in a 15% reduction in cloud storage costs and improving resource utilization across teams.
- Developed and deployed automated model monitoring workflows using AWS SageMaker Studio and Model Monitor APIs to detect model drift, track prediction quality, and streamline ML lifecycle management for future deployments.
- Conducted extensive testing of SageMaker containerized environments (CPU and GPU) across JupyterLab, Code Editor, and RStudio, ensuring compatibility and performance for various user profiles.
- Facilitated sprint planning, task management, and team communication using Jira, contributing to a cross-functional agile workflow.
- Actively participated in the SageMaker V2 migration initiative, providing internal documentation and cross-team support to ensure a smooth upgrade path and technical alignment.
Data Operations Intern
Madison Energy Infrastructure - Vienna, VA | June 2023 - August 2023
Team : Asset Management
- Developed an automated, serverless ETL pipeline using AWS Lambda to ingest, clean, and process monthly energy and financial data, generating structured Excel reports used in stakeholder briefings and audits.
- Redesigned the client-facing billing dashboard using Microsoft Power Apps, reducing load times from several minutes to seconds and improving the user experience for both internal teams and external clients.
- Integrated secure access controls into the dashboard to restrict visibility of sensitive financial data on a per-client basis, eliminating unauthorized access risks and enhancing compliance with data governance policies.
- Implemented automated backup and recovery workflows to ensure data availability and prevent reporting disruptions, strengthening the reliability of the reporting infrastructure.
Machine Learning Engineer
Qualcomm - Bangalore, KA, India | April 2021 - August 2022
- Built a data preprocessing pipeline to automate the cleaning and transformation of daily timing reports, enabling seamless data extraction and structured dataset creation for modeling workflows.
- Developed and deployed an LSTM-based classification model to automatically categorize extracted data, saving approximately 10 manual hours per week and improving operational efficiency.
- Implemented continuous model retraining processes to maintain classification accuracy over time, proactively addressing model drift and ensuring consistent long-term performance.