CS @ Georgia Tech | IoT Intern @ IBM
Fullstack developer interested in Reinforcement Learning
My journey involves learning and exploring the fundamentals behind Reinforcement Learning + Robotics. I'm equally fascinated by the nuances of full-stack development, continually exploring Angular, Vue, React and the intricate differences between frontend/backend frameworks.
When I'm not at the computer, you can find me experimenting with my camera, hitting the gym, or eating out!
Jan 2024 - Apr 2024
Helping maintain a robotic project of a Jetson Nano that moves around and detects hazardous materials in a room using computer vision. Another future project consists of building a GenAI chatbot through Google Vertex AI and Dialogflow.
May 2023 - Aug 2023
Built several frontend pages of LogicomUSA's call management software with Vue.js and AWS. The app is being currently used by Logicom's internal and external clients.
Jan 2023 - May 2023
Led a team of 4 developers, 2 designers alongside a product manager to create an app for the non-profit Southface making it easy to search, filter and compile sustainable building standards.
May 2022 - Aug 2022
Developed a frontend settings page for NCR's new Point-of-Sale software for restaurants through React.js. Also implemented Apple Pay, Google Pay and PayPal in the backend of the software using Nest.js and FreedomPay APIs.
Deep Q-Learning RL Agent Balances a Pole
Implemented a Deep-Q-Learning Reinforcement Learning Agent on OpenAI Gym's cartpole environment.
Monte Carlo RL Agent Plays Blackjack
Implemented the Monte Carlo Reinforcment Learning algorithm to play the game of Blackjack
Shoe Price Detection Using Computer Vision
Trained a custom ResNet and VGG image classification model to to detect the price of a shoe using scraped images from several sites including Nike, Adidas, Puma and more.
Robot Warehouse Automation
Created a robot to explore, locate and retreive items in a simulated warehouse using Frontier Planning for exploration, RRT for path planning and PID controllers for control.
A syllabus-to-calendar application that schedules calendar events based on lecture & exam information detected on the syllabus using NLP.
🏆 1st place Google Cloud's sponsor prize @ HackTX 2021
Made a full-stack web app of my 'fake' crypto coin to learn the ideas behind blockchain such as a elliptic hash functions, SHA256, Proof of Work and more.