Varsha Balaji

Computer Science Grad Student @ UIC & ML Researcher

Dedicated tech enthusiast excelling in ML and data science, orchestrating innovation through insightful data mastery.

About Me

Passionate about technology and innovation

I am currently pursuing my Master's in Computer Science at the University of Illinois Chicago, where I hold a CGPA of 4.0. My primary areas of interest include Machine Learning, Artificial Intelligence, and their practical applications across domains. My portfolio reflects a range of hands-on experience, peer-reviewed publications, and technical projects focused on solving real-world problems with measurable impact. My technical journey began with Natural Language Processing and has progressively expanded to encompass core machine learning and deep learning methodologies. I am currently involved in research exploring advanced neural architectures and am also exploring the emerging area of Agentic AI, with a focus on autonomous systems and task-driven agents. With a solid foundation in both theory and implementation, I am driven to build scalable, efficient, and ethically grounded solutions that address pressing challenges in today's data-driven world.

Experience

Professional and academic experience in data science, machine learning, and education

Professional Experience

Data Science Intern

Preludesys

February 2024 – May 2024 | Chennai, India

Conducted R&D on AI, specifically on beacon technology and cross-app marketing, researching the role of third-party cookies in extracting customer behavior from social media platforms.
Developed a working prototype for customer behavior analysis using Machine Learning models, designed for integration into the CRM platform on Dynamics 365.

Data Science Intern

Synergy Maritime

July 2022 – August 2022 | Chennai, India

Worked with the data science team on ML-driven virtual sensor in "Scrubber Management" using regression models.
Implemented advanced data preprocessing techniques and leveraged model hyperparameter tuning, resulting in a substantial accuracy improvement from 0.60 to 0.85, marking a 25% enhancement.

Academic Experience

Graduate Teaching Assistant – Capsim Coach

University of Illinois Chicago, Managerial Studies Department

January 2025 – Present | Chicago, IL

Assisted over 300 students across 12 sections of MGMT 495 and IDS 495 courses.
Delivered in-class presentations, facilitated project-based simulations, and provided one-on-one support to clarify concepts and address student queries.
Helped students navigate business strategy simulations by explaining financials, competitive positioning, and operational decision-making in a real-world business context.

Projects

Innovative solutions combining machine learning, data science, and software engineering

Facial Anti-Spoofing System

The project involves developing a facial anti-spoofing system powered by Vision Transformer architectures. By introducing cross-layer relation modeling and multi-scale feature fusion, it enhances robustness against spoofing attempts under varied lighting and pose conditions. Real-time deployment is achieved through a Flask-based interface with WebRTC-enabled client-server communication for live image validation.

Vision Transformers Computer Vision Flask WebRTC

Background Removal using PCA

In this project, Principal Component Analysis (PCA) is used as a dimensionality reduction tool for isolating foreground from background. By extracting essential components, PCA enhances clarity, focusing on main elements and reducing background noise for improved visual quality.

PCA Dimensionality Reduction Image Processing Machine Learning

Airline Reservation System

This project facilitates seamless airline reservations through real-time data streaming via Apache Kafka. Uses the analytical power of Apache Spark and Spark SQL to extract valuable insights from reservation data. Enhances efficiency and precision in managing airline bookings for an optimized customer experience.

Apache Kafka Apache Spark Real-time Streaming Data Analytics

Immunization Effectiveness in Rural Areas

This project employs machine learning regression models to analyze demographic, clinical, and genomic data, predicting childhood immunization likelihood among Indian women. It addresses regional variations, including urban, hilly, and rural areas, aiming to enhance vaccine coverage and personalize strategies for vulnerable populations.

Machine Learning Regression Models Healthcare Analytics Data Science

PennyWise - Expense Management System

PennyWise is a web application streamlining bill splitting. With a user-friendly interface, it allows easy receipt uploading, item selection, and precise cost calculation, offering an efficient tool for groups managing shared expenses.

Web Development JavaScript UI/UX Design Financial Tech

VoidVoices

This web application, designed for enhanced accessibility for differently-abled individuals, employs deep learning (OpenCV) for sign language to text conversion and JavaScript for Braille conversion. It addresses various aspects of disability, providing a user-friendly platform for improved communication and inclusivity.

Deep Learning OpenCV Accessibility Computer Vision

Publications

Research contributions in machine learning, natural language processing, and healthcare applications

Deep Learning-Based Renal Stone Detection: A Comprehensive Study and Performance Analysis

Authors: Varsha Balaji, Shahul Hameed T, Priyadharsini R

Journal of Statistics and Applications - ISSN 2454-7395

View PDF

CSE_SPEECH @LT-EDI-2023: Automatic Speech Recognition for Vulnerable Old-aged and Transgender People in Tamil

Authors: Varsha Balaji, Archana JP, Bharathi B

Proceedings of the LT-EDI 2023 – Third Workshop on Language Technology for Equality, Diversity and Inclusion, pages 204–208, Varna, Bulgaria, Sep 7, 2023.

View PDF

NLP_SSN_CSE @DravidianLangTech-RANLP 2023: Fake News Detection in Dravidian Languages using Transformer Models

Authors: Varsha Balaji, Shahul Hameed T, Bharathi B

Proceedings of the Third Workshop on Speech and Language Technologies for Dravidian Languages

View PDF

NLP_SSN_CSE at HOPE2023@IberLEF: Multilingual Hope Speech Detection using Machine Learning Algorithms

Authors: Varsha Balaji, Aishwarya Kannan, Aishwarya Balaji, Bharathi Bhagavath Singh

Proceedings of the Iberian Languages Evaluation Forum (IberLEF 2023)

View PDF

Let's Connect!

Get in touch with me at: varshabalaji04@gmail.com