Rohan's Portfolio

Education

University of Texas at Arlington

Master of Science, Computer Science

Specialization : Intelligent Systems, Cloud Computing and Databases

Aug 2023 - May 2025

Vemana Institute of Technology

Bachelor of Engineering, Information Science

Specialization : Artificial Intelligence, Machine Learning

Aug 2018 - July 2022

Experience

Graduate Teaching Assistant

University of Texas at Arlington - CSE Department

August 2024 - Present

Serving as a Teaching Assistant for CSE-6632-Cloud Computing and Big Data under Dr. Leonidas Fegaras and Dr. Zaman Noor. The role includes conducting weekly office hours for student assistance. It also includes procturing and grading the exams and assignments throughout the semester.

Summer Resident Assistant

University of Texas at Arlington - Upward Bound

June 2024 - July 2024

Part of the Upward Bound program which aims at providing upcoming university students with the experience of the university in the form of summer programs. I handled the students in terms of safety and their overall experience with events when they were at the dorm.

Data Science and Business Analytics Intern

The Sparks Foundation, Remote

July 2021 - August 2021

Learnt the foundations of data analysis in Python. Worked on the applications of how different libraries and tools can be used to visualize data. Built a model from the learnings that could predict the results of a student in an upcoming examination based off the study hours they put in before the exam.

Java Intern

Campus.Build, Bangalore, India

January 2020 - February 2020

It was an early program where I was selected as a Java Intern to work at the startup. I had the role to work the core team to be a part of the product they were developing and also understand the industry standards.

Projects

MegaCosm - Detection and Classification of Astronomical Objects

MegaCosm is a Machine Learning model designed to detect objects in an image in real-time using the YOLO algorithm and Convolution Neural Network.
In this case the model has been trained by annotating with astronomical objects like the comets, stars, asteroids and many more.
Once the raw image with these objects is fed to the model, it will be able to detect the object and draw a bounding box around it and also specify the accuracy with which it detected the object.

Subscription Classifier

Created a machine learning model to predict client subscription to a term deposit, managing a dataset of around 18,000 records. Conducted meticulous data cleaning and preprocessing using pandas to ensure accurate and complete datasets suitable for classification.
Utilized decision tree and random forest machine learning algorithms on the prepared data to evaluate and interpret the performance of each classifier, leveraging libraries such as sklearn, matplotlib, numpy, and seaborn.

Hadoop Mapreduce

Developed a project using the Java Programming language using distributed systems concept Map Reduce to perform matrix multiplication.
Here the idea is that data has been split and been fed in different systems but now has to be fetched and then processed to given back as the results to all the nodes in the system to maintain consistency.
It also involves the concept of big data and horizontal scaling.

Red Blue Nim

Red Blue Nim is an Artificial Intelligence Agent 2-Player game where one the of the players could be a human or AI. Each turn a players chooses to eliminate few marbles from the other players stack.
It is played in two version standard and misere.Depending on the version the person with last remaining marbles wins or loses points.
It uses the Alpha-Beta pruning with optimization to make the agent extremely intelligent and take decisions calculating all the possibilites.

Crime Record Management System

It is a standalone web application built using HTML, CSS, JavaScript, PHP and XAMPP to store all the details regarding a crime.
It has information from where the crime was committed to victim details all stored in a relational database. It using the MySQL as the database for it. We access it using the PHPmyadmin.
It also has advanced features like backup using triggers from SQL.

Space Trivia Application

Developed a mobile application in Android Studio using Java and XML offering extensive gameplay, multiple difficulty levels, and engaging features; enhanced user engagement by 9%.
In each difficulty level, the user is given few mutliple choice questions with a certain timer before which they have to choose an option. It uses card style for the user interface. The total score is displayed at the end after all the questions are answered.

Hello, I'm