Abhay Pai

Researched and implemented synthetic data generation techniques with privacy guarantees using TensorFlow's privacy framework as a part of my Master's Thesis in collaboration with Otto-von-Guericke University. The primary focus was to embed recently set differential privacy norm for training deep learning and machine learning model to generate artificial data that would effectively protect the users (or customers) sensitive information

Implemented emotion identification from advertising videos and further analyzed for digital marketing Contributed to extend and modify the existing script to identify logos from advertising videos Researched for implementing optical character recognition for extracting vehicle tires information (including DOT code) and documented challenges along with a specific approach to achieve the goal successfully

Implemented a basic recommendation engine in python using customer data that can be used for purchase related suggestions to customers Analyzed customers demographic data for categorizing them into multiple groups using machine learning. This information is then used by marketing and sales department Contributed to build a model for forecasting in-and-out payment cash flow within the Finance Team. Generated Tableau dashboards for monitoring different financial motives

Review test case deliverable created by team members Designed testing suite for specified functionalities Generated defect reports and co-ordinated within the development and testing teams to ensure timely action

Performed in-depth testing for critical document processing system to ensure high quality performance in production Participate in productive improvement initiatives within the team Contributed to an in-house test developed automation tool

Master Thesis: 'Synthetic data generation with privacy' Team Project 1: Implementation of differential privacy in K-Nearest Neighbors method for Sequential Data Team Project 2: Web Crawling, Indexing and Image Retrieval using Lucene