Experience
May 2023 - Aug 2023
Data Science Intern
USRA-NASA
Technologies:- Python, QGIS, Keras with Tensorflow, Deep learning.
- Achieved 96% IoU accuracy rate by adopting wide array of ML techniques and optimizing the U-Net architecture with a ResNet-34 backbone resulting in effective classification of the burned and unburned regions in satellite land images using semantic segmentation.
- Utilized QGIS software to produce labeled images for GeoTIFF files and integrated various data augmentation techniques to enhance data quality and precision, demonstrating a focus on problem solving to optimize model performance.
Jun 2021- May 2022
Data Analyst
Optum - UnitedHealth Group
Technologies:- SQL, Python (Numpy, Pandas), Tableau, MySQL.
- Saved 4-5 hours of daily manual effort and reduced claim sync errors by implementing the automation of Python batch jobs to streamline healthcare data extraction, transformation and loading from the DB2 server to the application server.
- Improved decision accuracy by 25% by developing automated statistical reports with Tableau for senior leadership and cross-functional teams. These reports highlighted KPIs to understand business dynamics and forecast demand, identifying high-rejection claim sites and optimizing data source monitoring.
- Formulated complex SQL queries in MySQL to enhance data quality within the data center supply chain, resulting in a 20% reduction in data defects and improved problem resolution efficiency by 30%.
May 2019 - Jul 2019
Machine Learning Intern
Ducat
Technologies:- Python (Pandas, Matplotlib, scikit-learn, Keras with Tensorflow), PowerBI, Excel.
- Increased crop yield potential by developing end-to-end models in Python, conducting Exploratory Data Analysis (EDA) and data manipulation on Indian soil data to identify nutrient deficiencies and optimize agricultural decisions.
- Achieved a significant 15% reduction in pesticide costs by designing dashboards with PowerBI to present insights, optimizing farming techniques and pesticide use.