About me
I am currently pursuing Masters of Science in Computer Science from Stony Brook University in New York, United States.
As part of my thesis project, I am currently working in SBU CV Lab on building a assistive robotic arm to make self feeding task possible for people who cant physically feed themselves.
I have three years of industry experience working in Visa Inc. as a Software Developer. My interests include Distributed Systems, Machine Learning and Computer Vision.
Education
Stony Brook University, New York, US
GPA : 3.84/4.00
Relevant Courses : Machine Learning, Introduction to Computer Vision, Probability & Statistics for Data Science
Teaching Assistant : CSE 114 : Introduction to Object Oriented Programming under Prof. Praveen Tripathi.
Motilal Nehru National Institute of Technology, Allahabad, India
GPA : 8.25/10
Relevant Courses : Data Structures, Object Oriented Modeling, Databases, Software Engineering, Operating Systems & Networking
Undergraduate Research : Worked under supervision of Prof. Rupesh Dewang on Sentiment Analysis to classify product reviews collected from twitter.
Work Experience
Visa Inc. - India (Senior Software Developer)
-
Data Lake Controller| Java, Spring Boot, Kafka, Docker, Airflow
- Designed micro-services for Data Lake Controller which provided a metadata driven implementation to manage components of Visa Data Platform automating 68% of ETL pipelines.
- Implemented services to auto-generate data pipelines and deploy (avg. time of 2 secs) on distributed schedulers.
- Delivered scheduler and checkpoint micro-services to deploy and run data pipelines (700+) on apache airflow. The service allows a user to author, schedule and monitor the data flow and view the dataset lineages.
- Integrated Data Lake with ML Platform allowing data scientists to build, deploy and monitor ML models efficiently.
-
Data Quality Platform| Java, Spring Boot, MySQL, HDFS, ETL, Grafana
- Developed a data management and analytics platform for automating data quality checks across 28 heterogeneous data environments.
- Scaled it to perform more than 50,000 checks across completeness, consistency, and timeliness dimensions of data.
- Spring Boot application allowing users to configure rules, monitor them and generate alerts as part of ETL pipelines.
- Cron based scheduler to execute data quality checks based on the frequency of checks.
-
Root Cause Analysis| Python, Time series analysis, Cross-correlation
- Formulated a root cause analysis framework to correlate event trends and detect root cause of a quality issue with 72% accuracy.
- Built an event repository of multiple internal and external events affecting a dataset.
- Built a correlation engine to correlate events with an anomaly using a probabilistic model.
- Filed a US Patent for the same - PCT/US2018/066540
Visa Inc. - India (Summer Intern)
-
Forecasting and Anomaly Detection| Java, Spring Boot, Holt Winters, Predictive Analysis, Outlier Detection
- Implemented anomaly detection framework using Holt Winter's triple exponential smoothing algorithm.
- Spring Boot application to run the detection algorithm on event of new transaction. Data was consumed from HBase and anomalies were written to mysql for root cause analysis and alerts.
- Weekly, monthly, yearly, holidays and special events as trends for better accuracy. It helped to reduce the percentage of false alerts by 62%.
- Generalized the framework to incorporate other forecasting models like Facebook's Prophet, Moving Average models.
