I am an UMass Amherst CS grad student with 3 years of experience in data engineering. Passionate about tech innovation and AI's transformative role.
Implemented a scalable, fault-tolerant data ingestion pipeline utilizing HDFS, Spark, and Scala for distributed processing and data manipulation, handling 15 million daily transactions.
Identified and resolved a data discrepancy across data centers, preserving $240k in revenue through root cause analysis, data lineage exploration, and code review
Created a modular, high-performance ETL tool with Python and Hive to automate data extraction and transformation across applications, achieving a 43% reduction in code redundancy and a 10% decrease in resource utilization.
Spearheaded the implementation of a scalable, config-driven utility for automating data pipeline testing, resulting in a 32% reduction in manual testing workload.
Successfully remediated 3 years of data inconsistencies involving 10 billion records through data lineage analysis, business logic refactoring, incremental data backfilling, and continuous monitoring to avoid disruption to end users.
Built a full-stack forecasting application with React, Django, and Facebookâs Prophet for predicting transactional data, achieving an 18% accuracy improvement in anomaly detection.
Developed a multiscale CNN with 6 convolutional layers to extract and integrate local and global features, employing random forest classification for high-accuracy outcomes exceeding 96% across diverse datasets.
Innovated a novel approach utilizing contrast enhancement and pixel grouping to extract RPE layer and baseline from retinal OCT images for AMD diagnosis, achieving a 96.66% accuracy on a dataset of 2130 images.
Led a 5-person team in developing a Python and Tkinter-based chess GUI with reverse rules, incorporating the Min Max algorithm, resulting in an AI opponent achieving an 80% win rate.
GithubDesigned a React and Django-based full-stack web app, employing RESTful APIs for a single-click Job-Board platform.
GithubImplemented the Prioritized Sweeping algorithm in Python for Blackjack, resulting in a 70% win rate across 100 gameplays.
Github