Federal Home Loan Mortgage Corporation (Freddie Mac)
McLean, VA, USA
Built a GraphQL API in Python using FastAPI, Strawberry and SQLAlchemy to connect multiple microservices, reduced latency by 60%
Automated performance testing on pre-prod environments with Python script, saving 30+ developer minutes each deployment using Jenkins CI, Docker and Kubernetes on RedHat OCP
Wrote a BASH CLI to evaluate pytest suite differences between two GIT SHAs, improved coverage
Implemented vectorized regex for transaction processing and matching, improving performance by 10% in 99th percentile
Implemented asynchronous database commit functionality, enhancing system efficiency, reducing CPU utilization across multiple K8s pods by 30% on average
2022-04-11|2023-03-20
Programmer Analyst
Amazon
Minneapolis, MN, USA
Built a CI/CD service to customize freight labels, retaining $1M+ business while expanding in EU, led architecture design, implementation and monitoring using AWS
Built RESTful APIs to connect the service to a 2-phase UI with React JS for the frontend and Java and Spring for the backend
Wrote Spock, JUnit and Mockito unit-tests and integration-tests to reduce instances of bugs, participated in design and code reviews with senior engineers
2021-11-015|2022-04-08
Data Analytics Engineer
Open Systems International
Medina, MN, USA
Reduced SQL query runtime for an IEEE KPI metrics query by 99% from 50 minutes to under 25 seconds by analyzing query plans and optimizing using custom indexing
Led migration and conversion efforts from legacy database schema to new schema for 100M+ records, automated performance testing for SQL queries using custom UDFs with parametric queries
2019-08-14|2021-05-07
Master of Science - Data Science
Indiana University
Bloomington, IN, USA
Coursework focused on data structures, algorithms, databases and machine learning
2018-06-07|2019-06-28
Associate Data Scientist
AlgoAnalytics
Pune, MH, India
Generated revenue of $50K, built and optimized a custom YOLOv3 object detection model to automate the detection of spray paint defects in images of manufactured parts with over 89% accuracy
Automated synthetic data generation using Blender’s Python API and GANs, reduced KL divergence between synthetic and real data by 77%, improved object detection on a subset of Cityscapes by an average of 11% IoU across 7 categories
Led client meetings to understand requirements and modelled solution pipelines, implemented research papers, created documentation, and deployed solutions as Flask web applications
2014|2018
Bachelor of Technology - Computer Science and Engineering
Government College of Engineering
Amravati, MH, India
Coursework focused on data structures, algorithms, databases and mathematics and other topics