Vien Tran

About me

An aspiring student from Georgia Tech willing to improve his endeavors in all realms. Be it the physical realm of combat competing against monstrous athletes' or learning the intricate world of technology and engineering, the endless amount of potential is always unraveling till the end.

My work blends embedded systems, machine learning, and software engineering with a focus on real-world impact.

As the founder of Strike, I developed a smart boxing tracker that captures real-time performance data using embedded sensors and Bluetooth connectivity, enabling safer and more informed training for athletes. Through my research with the Structured Information for Precision Neuroengineering Lab , I contributed to human-in-the-loop AI systems by designing web-based interfaces that streamline data collection for sentiment analysis and natural language tasks. This experience sharpened my understanding of human-centered machine learning and scalable experiment design.
At the Najam Institute, I also apply my technical skills to community development, serving as a Content Director and Technical Editor focused on Islamic education and outreach. Every experience has reinforced my belief that technology is most powerful when it improves lives — and that's the mission that drives me.

Education

Experience

Developed a smart boxing tracking device that measures key metrics like power, speed, and force, and transmits data in real-time via Bluetooth to a smartphone app. Enhanced the accuracy of impact measurements up to 109 Pa through refined product testing, ensuring both device durability and user comfort. Led product launch and marketing efforts, positioning the device as a valuable tool for professional training in boxing gyms, while also promoting its appeal to individual fitness enthusiasts looking to improve their performance.

Designed and deployed a web-based experimental interface to collect user responses for sentiment classification tasks, enabling high-quality data labeling at scale. Integrated with the Prolific platform to streamline participant flow and manage over 100 contributors efficiently. Contributed to active learning strategies that reduced human input without sacrificing model accuracy, enhancing the system’s ability to learn from minimal supervision and improving overall annotation speed.

During my internship as an Analyst Lead for RTC Information Systems at OMOMI Industries / OLTS, I enhanced operational efficiency and supported strategic decision-making by coordinating transportation logistics, analyzing data, and generating optimization reports, which improved delivery times and reduced costs.

Refined and optimized hardware for delivery drones capable of picking up and delivering packages on campus. Led troubleshooting and repair of electrical components and circuits, improving the drone's performance, durability, modularity, and reliability. Additionally, contributed to the development of an improved drone carrier design to enhance delivery efficiency.

Projects

Developed an embedded telecontrol glove system using C++, Arduino, and an ESP-32 to enable real-time VR interaction with a 3D-printed robotic hand. Finger motion was measured via five potentiometers linked by tension strings, calibrated through a custom-built button interface for precise tracking. Integrated tactile feedback by mapping five feedback buttons on the robotic hand to corresponding servos on the glove, simulating the sensation of physical contact. The system created an immersive, bidirectional control loop with potential applications in assistive tech, robotics, and virtual environments.

The project involves developing an AI chatbot that has progressed through several levels of complexity. Starting with a basic chatbot structure and input handling, the project evolved by integrating OpenAI's GPT-4 model for intelligent responses, deploying the web app on AWS EC2, and optimizing it using AWS Bedrock and RAG (Retrieval-Augmented Generation) for knowledge-based responses. The final implementation leverages various AWS services, Next.js, and Pinecone for efficient data handling and vectorization, resulting in a sophisticated, scalable chatbot capable of intelligent, context-aware interactions.

• Video DEMO

Designed and implemented a five-stage pipelined RISC-V processor in SystemVerilog, supporting instruction-level parallelism and modular microarchitecture. Boosted execution throughput by 6.44× through hazard mitigation techniques such as pipeline stalling, data forwarding, and branch prediction. Validated design correctness using ModelSim, performing detailed waveform analysis, timing verification, and register tracking to ensure accurate instruction flow and reliable processor performance.

Designed and implemented an FPGA-based servo controller using VHDL and Assembly to drive five independent servos concurrently, increasing hardware scalability by 400%. Developed precise timing logic and documented system architecture, including signal constraints and interfacing methodology. Addressed challenges in concurrent control and synchronization, ensuring reliable pulse-width modulation and consistent servo behavior under real-time conditions.

Engineered a real-time space shooter game using C on a custom Arduino-based embedded platform, featuring dynamic UI elements and event-driven input for an interactive user experience. Focused on optimizing system performance by restructuring the game loop with memory-efficient data handling and loop unrolling, resulting in a 50% reduction in execution latency. The project demonstrated embedded-level game logic execution and responsive hardware integration under tight timing constraints.

• Video DEMO

Developed a responsive web application to track pantry inventory in real time, enabling users to monitor food supplies, log usage, and receive stock alerts. Implemented Firebase for backend data storage, real-time database sync, and user authentication, ensuring seamless access across devices. Deployed the full-stack app using GitHub Pages, integrating version control, CI/CD, and collaborative development workflows. Designed intuitive UI components to enhance usability and encourage sustainable household management.

Developed a comprehensive MIPS project that performs a binary search on DNA sequences, focusing on efficiency by reducing the dynamic instruction count by 50% and minimizing the static instruction count by 41%.

Resume

Links

• Github
• LinkedIn

contacts

Element

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';