Jay Borkar Portfolio

Architected Data Analytics Platform for Solution Engineering (presales) organization to ingest, analyze and visualize the data.
Developed Analytics tools and Tableau Dashboards to optimize tracking and reporting across Solution Engineering organization.
Enabled SE (presales) organization in defining KPIs, Success metrics to measure business performance and drive scale.
Fostered a culture of using data to inform product and business strategies across Solution Engineering organization.
Designed and built efficient Data pipelines to ingest, model, transform and visualize the data across SE business portfolio.
Standardized data strategy, which is focused on insights to gain a better understanding of SE users’ behaviors, and how that directly or indirectly impacts sales.
Partnered with our stakeholders’ team to provide actionable data insights to SE business leaders.

Coordinated engineering scrums, sprint planning, technical discussion.
Designed, developed Airflow DAG ETL Pipelines that schedules, triggers, monitors ETL Jars/ Python Wheel to ingest data to the Data Lake.
Built, deployed Streaming pipelines using Spark and Kafka for real-time streaming to make better, faster decisions.
Built and deployed Machine Learning pipelines that uses Sales, Customer and Geospatial data to assists the business to predict realistic market potential for a given census tract and core customer base (trade/catchment areas) for all the Footlocker stores in the US.
Scaled and Optimized the Machine Learning project lifecycle using PySpark, MLflow, Airflow DAG, Jenkins CI/CD pipeline.
Collaborated with Principal Data Scientists to improve the ML model speed and reduce performance time by 90% using PySpark.
Researched, Analyzed and tested open-source suite of tools that enables large-scale geospatial querying and analytics on distributed computing systems.

Designed and Implemented Data Exploration routines to do statistical analysis and visualization.
Built Classification models such as Logistic regression, SVM, Random Forest, Gradient Boosting Machine (GBM), Deep learning to predict whether or not an applicant is capable of repaying a loan.
Predicted Risk by predicting Loan Defaulters, Evaluated Classification models by Accuracy, Confusion Matrix, Precision, Recall, True Negative Rate (TNR), False Discovery Rate (FDR), Gain Chart, Lift Chart, K-S Chart, ROC - AUC chart.
Deployed the Final solution as a Web application (Restful API), Link for demo : default-risk.herokuapp.com

Analyzed business needs, designed and developed automated tools.
Automated Product Book Titles search task by developing a proof of concept tool that uses Machine learning/Natural Language Processing to perform Fuzzy String Matching on Book Title data to return probable Book Titles matches with a confidence score.
Automated Data cleaning by developing a POC tool that uses Machine learning to predict duplicate records in the database.
Developed the solutions in an agile environment on Wiley Data Science Environment that contains JupyterLab and Cloud compute machine, Implemented solutions in Python with Dedupe, Fuzzywuzzy libraries on database having more than 450,000 Book titles.

Appointed as Teaching Assistant for CS405-Computer, AI and Society course, Worked with Prof. Casimir Kulikowski. Taught material, Reviewed and graded weekly assignments, papers and exams.
Effectively and accurately communicated with Professor Kulikowski and students regarding grades and potential areas of improvement.
Engaged in student discussions and took notes for weekly debrief, Kept track of attendance and student participation for a class of 40+.