Garegin Mazmanyan

About Me

I'm a computer science graduate student at the University of Arizona, specializing in the intersection of full-stack development and machine learning. My passion lies in building scalable, intelligent systems that solve real-world problems.

                            
                            MS in Computer Science
University of Arizona

                            
                            AI Robotics Research
University of Arizona

                            
                            AWS Certified
3 Professional Certifications

Professional Timeline

2025 - Present

Engineering Robotics Lab

University of Arizona

Leading autonomous driving research that integrates computer vision with natural language processing to develop intelligent vehicle systems. My current project focuses on creating multi-modal autonomous agents capable of interpreting spoken commands while providing real-time explanations of driving decisions. This research advances the field by bridging the gap between mechanical automation and truly intelligent robotic systems.

Python C/C++ Computer Vision Deep Learning TensorFlow PyTorch ROS2 AI Research

2024 - Expected 2026

Master of Science in Computer Science

University of Arizona

Pursuing an accelerated master's program that enables concurrent completion of graduate coursework while finishing undergraduate requirements. Specializing in artificial intelligence and machine learning with emphasis on practical applications in autonomous systems. The program provides advanced theoretical foundation combined with hands-on research experience in cutting-edge robotics and intelligent systems development.

Machine Learning Deep Learning Artificial Intelligence Computer Vision NLP Research Methods Advanced Algorithms Data Structures

2024 - 2025

Software Engineer Intern

AEYESAFE

Developing and implementing scalable cloud infrastructure solutions while gaining comprehensive industry experience in modern software development practices. Successfully migrated legacy applications to AWS cloud architecture, utilizing containerization technologies and automated CI/CD pipelines. Achieved 50% reduction in deployment time through strategic implementation of Docker-based workflows and collaborated effectively with cross-functional teams to deliver robust software solutions.

Docker AWS ECS Fargate DynamoDB Lambda API Gateway CI/CD Python YAML

2021 - 2025

Bachelor of Science in Computer Science

University of Arizona

Established comprehensive foundation in computer science fundamentals while developing leadership experience through teaching roles. Served as Undergraduate Teaching Assistant and Course Coordinator, mentoring over 300 students in software development and programming concepts. Earned multiple professional certifications including AWS Cloud Practitioner, AWS Developer Associate, and AWS Machine Learning Specialty. Successfully balanced academic excellence with practical skill development in cloud technologies and modern software engineering practices.

Python Java JavaScript React Node.js HTML/CSS SQL MongoDB Git AWS Certified Teaching

Featured Projects

Decentralized Affine Transformation for Drone Swarms

Completed

Experimental validation of decentralized affine transformation in multi-agent systems using teams of mini-quadcopters. This research enables drone swarms to safely navigate constrained, obstacle-rich environments while allowing aggressive changes in inter-agent distances through a leader-follower framework. Published research demonstrating asymptotic convergence and safe multi-agent coordination.

🔹 Decentralized control 🔹 Leader-follower coordination 🔹 Obstacle avoidance 🔹 Experimental validation

ROS2 Python Decentralized AT Crazyflie 2.1 Motion Capture Research Publication

VR-Based Control of Multi-Copter Operation

Completed

Immersive VR teleoperation system for multirotor flight featuring third-person view with real-time 3D reconstruction. Integrates NVIDIA Jetson Orin NX, Meta Quest 3, and ROS2-WebXR for low-latency control in unmapped environments. Experimental validation demonstrated 32% improvement in obstacle clearance and zero collisions compared to traditional first-person view systems.

🔹 Third-person VR perspective 🔹 Real-time 3D SLAM 🔹 Low-latency control (16ms) 🔹 Enhanced spatial awareness

Virtual Reality ROS2 WebXR Meta Quest 3 NVIDIA Jetson Three.js Real-Time SLAM Research Publication

CSC110 Coding Platform

Completed

A comprehensive educational platform revolutionizing computer science education. Features include an interactive code editor with syntax highlighting, real-time Python execution environment, automated test validation, and progress tracking. Increased student engagement by 40% and improved learning outcomes through immediate feedback mechanisms.

🔹 Real-time code execution 🔹 Automated testing 🔹 Progress analytics 🔹 Interactive learning

React.js Node.js Express.js PostgreSQL JWT Docker

QCar2 Autonomous Driving with Natural Language

In Progress

Vision-Language-Action model adapting SimLingo from CARLA to QCar2 platform. Integrates InternVL2-1B for autonomous driving with natural language command processing and real-time explanation generation. Features GPU-accelerated inference, complete control pipeline from camera to vehicle actuation, and interactive commentary window for high-level commands.

🔹 Vision-Language-Action model 🔹 Natural language control 🔹 Real-time commentary 🔹 GPU-accelerated inference

Python PyTorch InternVL2-1B ROS2 QLabs CUDA Computer Vision NLP

AI-Powered Portfolio

In Progress

Professional portfolio website featuring timeline design and integrated AI chatbot assistant. Built with modern web technologies, includes smooth animations, responsive design, and intelligent conversation capabilities. Showcases career progression and technical expertise through interactive elements.

🔹 AI chatbot integration 🔹 Responsive design 🔹 Smooth animations 🔹 SEO optimized

HTML5 CSS3 JavaScript Particles.js AI API Progressive Web App