About Deepthi D’Souza
-
Viewed 391
About me
Experienced Data Scientist with a demonstrated history of working in Healthcare and Insurance industries. Skilled in Advanced Machine Learning, Deep Learning, Data Engineering, Big Data, Data Analytics, Data Intensive Computing on Cloud,Python programming and Database.
Education
-
- 2023
University at Buffalo, The State University of New York
Master of Science - Data Science
STEM Designated, GPA: 3.7/4.0 Coursework: Machine Learning, Deep Learning, Data Mining, Data Intensive Computing, Data Analytics, Statistics, Database
-
- 2018
Jawaharlal Nehru National College of Engineering (VTU), India
Bachelor of Engineering, Electrical and Electronics
First Class with Distinction, GPA: 3.8/4.0
Experience
-
2021 - 2022
Agilon Health
Data Scientist/ Machine Learning Engineer
• Designed and developed end to end ML lifecycle for ML models which predict Customer Churn on Databricks using MLOps.
• Utilized TensorFlow and Scikit-learn to build predictive models like Decision Trees, Random Forest, XG Boost., reducing churn rates and
achieving 85% model accuracy.
• Conducted A/B Tests, Hypothesis Tests, Unit and Integration Tests to build resilient ML systems.
• Implemented CI CD pipelines and automated model retraining which resulted in high accuracy and availability of models.
• Designed and developed a data platform for Patient EMR data using AWS services (EC2, Snowflake, Airflow).
• Collaborated with cross-functional teams, including sales and marketing, to communicate insights and recommendations. -
2020 - 2021
Accenture
Data Engineer
• Designed and implemented a real-time data pipeline to process raw data by integrating 100 million records from more than 15 data sources
using PySpark on AWS Databricks.
• Designed and implemented changed data capture using Delta Tables on cloud for 10million records every month which improved the changed
data update performance by 45%.
• Led the migration of SAS code to Spark, data from SQL server to S3 resulting in 50% lesser processing time after optimization.
• Collaborated with Data Scientists and Research and Development team in understanding the data requirement for ML models which predicts
customer retention for various insurance policies.
• Obtained meaningful insights, generated reports for business users using MS SSRS, PowerBI. -
2018 - 2020
Infosys Limited
ETL Developer
• Developed RESTful web services using Flask for post processing data from intelligent sources like OCR.
• Developed data capture process to get required data from defined positions from posters/image using OpenCV2.
• Designed and developed batch processing management on cloud Using Apache Nifi.
• Collaborated with clients from across the globe, gathering requirements and engaging with multiple stakeholders to develop and maintain 12
different data-pipelines.
• Designed and Developed Data warehouse and Built ETL pipelines to process and load 5 million records into Data Lake for downstream use.