Worked as a part of the product development team enabling Epic software to communicate clinical data with external systems. Responsibilities include:• Analysis, research, development and implementation of Epic software at healthcare institutions across USA.• Identifying ways to improve key healthcare delivery workflows.• Design and develop high quality patient critical systems integrated with the rest of Epic.• Code review of peer projects.• Frequent customer interaction and management of multiple projects with competing priorities.

Conducted research on the privacy guarantees and utilities of different data privacy techniques.Outcomes of my Masters thesis research under Dr. Daniel Kifer:Topic: Constructing Improved Differentially Private Classifiers - Derived an empirical Bayes’ estimator for estimating the mean of a multivariate double exponential (Laplace) distribution. 
- The estimator outperforms the maximum likelihood estimator and is the first of its kind for the multivariate Laplace distribution. 
- Designed and implemented a differentially private classification algorithm based on Naïve-Bayes’ formulation and showed that its performance can be improved without the use of any auxiliary information by using the derived estimator. 
Topic: Attacking Anatomy Privatization Scheme by Extracting Sensitive Information 
- Designed a Naïve Bayes’ classifier using Expectation maximization algorithm for the purpose of predicting the sensitive attributes of Anatomy sanitized data sets. 
- Described a method to obtain well-calibrated probabilities of prediction. 
- Compared to previous attacks, this attack algorithm is computationally simpler and gives comparable accuracy levels. 
Dissertation attached:

Learned functionality of analog and digital microwave devices like the Toshiba 7D6, 7D15, 7D20, Harris equipment, PunCom Digital equipment and other inline systems that required timely resolution and de-bugging in a schedule-sensitive scenario within the physical parameters of the numerous telecom towers.