Kavin Chandrasekaran

Data Science Intern

Clean Crop Technologies - (May 2022 to August 2022)

Architected an end-to-end data science framework that streamlined the adaptive design of experiments for sustainable food production research, minimizing manual input and accelerating research cycles

Implemented optimization models based on Gaussian processes to determine optimal in-lab experiment parameters, maximizing crop growth while preserving quality

Presented findings to senior management, resulting in adoption of new methodologies across research teams and reducing research cycle time

Research Assistant

Prof. Emmanuel Agu

Worcester Polytechnic Institute - (May 2018 to December 2024)

Pioneered a high-performance deep learning architecture using PyTorch, achieving 97% accuracy in recognizing and classifying ambulatory activities and transitions from smartphone sensor data, leading to more accurate patient mobility monitoring

Innovated a novel feature generation mechanism using NLP techniques, that improved complex human activity recognition by 6-23% in F1-scores, outperforming state-of-the-art and baseline models

Led a research project leveraging language models to enhance complex activity recognition performance, with the goal of enhancing patient monitoring, predicting ailments, and improving elder care

Sr. infrastructure Services Engineer

National Government Services - (July 2013 to December 2015)

Managed and monitored critical IT infrastructure, leveraging data analysis to optimize resource allocation, proactively identify performance bottlenecks, and ensure high availability

Developed and maintained monitoring dashboards and reports, utilizing data visualization tools to communicate key performance indicators

Pioneered an effort to integrate OEM into daily monitoring workflow and created a centralized dashboard for monitoring and troubleshooting the application using the product iDashboards

Responsible for creating and maintaining scripts to monitor the performance and availability of the application and create automated reports using data from Patrol, OEM and BAC (Business Availability Center)

IT Intern

National Government Services - (January 2013 to April 2013)

Worked on a CSR Staffing analysis to predict the CSR staffing requirements, based on previously known staff availability, requirement and their efficiency

Junior Support Engineer Intern

LikeMinds Consulting Inc. - (May 2012 to August 2012)

Provided support for AARP\u2019s identity management system and security infrastructure

Was responsible for ensuring application availability during scheduled maintenances, handling network failures, disk space issues, replication issues

Wrote a plugin which allowed the users to authenticate them via. Kerberos when the Kerberos feature is available and if not, it would automatically failover to the NTLM based authentication

Associate Instructor

Indiana University - (March 2012 to May 2012)

The course revolved around the economics of security and influence of security in business decisions and its impact in organizations. Was responsible for organizing the group discussions on various topics, assisting in preparing the exam and grading

Research Mentor

Indiana University - (January 2012 to May 2012)

Supervised a team of four undergraduate students through a research project called Facebook FakeFinder

The research was focused on finding fake profiles in Facebook using different metrics like the number of friends, relation with the friends, number of photos, tags on photos, privacy settings of photos, etc.

Developed a machine learning model that achieved 80% accuracy in detecting fake profiles. Received positive evaluations from the students and the faculty advisor for my mentoring skills