About Me

Welcome to my website! πŸ‘‹

I'm an aspiring software engineer with a passion for technology. My interests span machine learning, front-end and back-end development, as well as game development. πŸ§‘β€πŸ’»

In my free time, I enjoy playing basketball πŸ€, listening to music 🎡, cooking 🍳, and sleeping 😴.

My Projects

AI-Powered Financial Forecasting and Visualization Dashboard

    Made with:

  • Created responsive visualizations with MUI, Recharts, and React, showcasing interactive charts and tables for key financial metrics.
  • Built and maintained a backend server with Node.js and Express.js, efficiently handling API requests with a TypeScript-based, scalable architecture.
  • Handled over 4,000 data entries using Mongoose and deployed the application seamlessly with Vercel, Docker, and Fly.io.
  • Implemented a linear regression model using regression.js to forecast annual revenue based on historical data.

AI-Powered Auto-Updating Blog

    Made with:

  • Developed using Next.js for full-stack capabilities, with TypeScript ensuring type safety and maintainable code.
  • Integrated OpenAI APIs for AI-driven content generation and updates, enabling users to edit articles with AI assistance seamlessly.
  • Implemented Prisma and Neon for a robust, scalable database solution supporting efficient data management.
  • Styled the platform with TailwindCSS for a modern, responsive UI as well as an intuitive user experience and deployed on Vercel.

Chat Room with ChatGPT API

    Made with:

  • Implemented RTK Query for efficient API integration and utilized React Router for seamless navigation within the application.
  • Integrated Chat Engine to build a dynamic chat application interface and employed Redux Toolkit for robust state management across the platform.
  • Leveraged OpenAI’s technology to integrate advanced AI capabilities into the chat functionality and incorporated Heroicons to enhance the user interface with visually appealing icons.
  • Utilized Express.js as the backend framework and Node.js as the runtime environment to ensure efficient processing and scalable application performance.

SaaS Template Application

    Made with:

  • The SaaS template application is developed using Django for the backend, TailwindCSS for the frontend, Neon for database management, and deployed on Railway.
  • Payment integration is facilitated by the Stripe API, supporting both one-time and subscription payments.
  • Docker is used for containerization, with GitHub Action Workflows automating the CI/CD pipeline for testing and deployment.
  • The application structure includes checkouts, subscriptions, dashboard, customers, profiles, and a landing page for comprehensive service management.

Machine Learning for Algorithmic Trading

    Made with:

  • Developed algorithmic trading strategies using technical indicators like Bollinger Bands, RSI, and MACD to identify opportunities and enhance portfolio performance.
  • Utilized Jupyter Notebooks for iterative strategy development, data visualization, and machine learning integration in quantitative analysis.
  • Executed time-series analysis and rolling regression to apply advanced machine learning models for predictive analytics and trade optimization.
  • Performed dynamic parameter adjustments in trading algorithms based on real-time market data scraped from web sources to improve decision-making and responsiveness.

Moros - 2D story driven puzzle adventure game

    Made with:

  • Developed a game in a team of 8 (4 programmers) within one month for the GitHub & itch.io Game Jam, ranking #246 out of 5,000+ participants.
  • Collaborated in weekly team meetings to design the theme and style for Moros, a single-player 2D story-driven puzzle adventure, while tracking progress and refining mechanics.
  • Managed input systems (player movement, interactions, special abilities), ensuring responsive gameplay and integration with puzzle elements.
  • Developed dialogue systems and implemented player abilities like hiding during boss fights, adding strategic layers to gameplay.
  • Optimized performance by focusing on input responsiveness, frame rates, and puzzle mechanics, ensuring smooth gameplay during boss battles and complex sequences.

Kitchen Chaos - Single Player 3D game

    Made with:

  • Designed multiple game themes, including the main menu, loading screen, and gameplay, ensuring seamless transitions and a smooth user experience.
  • Created immersive kitchen environments with interactive elements that dynamically respond to player actions, enhancing interactivity while maintaining clean code structure and performance.
  • Developed responsive UI elements, including a tutorial interface, game pause, game over, delivery results, and menu notifications etc.
  • Integrated support for various game controllers, enabling responsive gameplay and customizable controls for enhanced player engagement and accessibility, built with Object-Oriented Programming (OOP) principles.

Amazon Clone

    Made with:

  • Enhanced transaction experience by calculating delivery dates with day.js and streamlined cart functions using OOP principles.
  • Adapted backend structures for comprehensive product data and ensured reliability with automatic tests using Jasmine.
  • Improved cart code modularity with JavaScript modules and optimized checkout.js using Async/Await for better performance.

Technical Skills

My Experiences

Software Developer & AI Engineer

Inspur Group Co., Ltd.

September 2024 - December 2024

  • Collaborated on the development of a multi-agent platform designed to support advanced Agentic Workflows.
  • Optimized input segmentation and retrieval within the RAG pipeline, improving system accuracy, reducing error rates by 50%, and enhancing overall user satisfaction.
  • Developed a document conversion pipeline utilizing Optical Character Recognition (OCR) to extract text, images, and tables from diverse formats (e.g., PPTX, DOCX, PDF, etc.) and converting them into Markdown.
  • Built a FastAPI service to generate custom Word documents using plain text or templates with placeholder replacement, integrating MinIO for efficient storage and retrieval workflows.
  • Used LabelImg to label images and organize training data for computer vision models.

Cross-Border Trade Assistant

Jiangsu Baiheng Energy Saving Technology Co., Ltd.

July 2023 - September 2023

  • Acquired skills in Shopify website development and successfully built an e-commerce site
  • Traveled to Shenzhen for an AI training business trip and obtained certification
  • Created a summary document on training content and facilitated an educational session for colleagues
  • Collaborated with organizers to coordinate a training session as the assistant to the CEO

My Education

2023 - 2028

University of Waterloo - Bachelor of Mathematics

Courses: Intro to Optimization, Intro to Financial Mathematics, Statistics, Business Finance

  • Major in Computational Mathematics
  • Major in Combinatorics and Optimization
  • Minor in Computing (Computer Science)
  • Enroll in the co-op program

Contact Me

Email

m69liu@uwaterloo.ca