- 🏒 SmartScore - Predict NHL Goalscorers
- AI application designed for Tim Horton's Hockey Challenge, predicts players probability of scoring in a NHL game
- Used machine learning with TensorFlow, pandas and NumPy to build an AI model using multiple linear regression
- Enhanced accuracy significantly by optimizing the AI model's metrics, elevating pick rate from 26% to 34%
- Actively implementing AWS Lambda functions to optimize serverless data collection and distribution
- Developed a Restful API to seamlessly integrate with full-stack applications, enhancing user experience
- Utilizes Draftkings API to provide precise recommendations for strategic betting decisions based on real-time data
- 🧠 Imaginate
- Co-developed a full-stack image guessing game using React and Flask, aimed at helping users learn to spot differences between AI-generated and authentic images based on a daily theme
- Built a text-to-image generation pipeline leveraging Hugging Face Diffusers and PyTorch for creating lifelike images, automating prompt generation with LLaMA 3.1, and integrating MongoDB for streamlined data handling and storage
- Constructed a CI/CD pipeline with GitHub Actions and AWS CloudFormation to automatically build, test, and deploy AWS Lambda functions, reducing development cycle time by 50%
- 🔠 Wordle Solving Bot
- Constructed a bot that analyzes a dictionary of possible words to solve the popular online game Wordle
- Implemented a logical algorithm that formulates guesses crafted to eliminate the maximum number of alternatives
- Created a comparison algorithm that dissects the ideal words and selects the word with the greatest popularity
- 🌐 Portfolio Website
- Built a responsive and dynamic portfolio website using Next.js, focusing on server-side rendering and seamless navigation.
- Leveraged npm for efficient package management, enhancing the project’s maintainability
- Utilized React JavaScript library to build dynamic and high-performing user interface components
