Mechanical Engineering 2026 + PEY | Minor in AI/ML & Engineering Business

About Me

I'm Aaryaman Singh, a third-year Mechanical Engineering student at the University of Toronto with a strong passion for engineering entrepreneurship and cutting-edge technology. Based in Toronto, I have a deep interest in startups, AI, and machine learning, and I thrive on curiosity and innovation. If you're looking for a driven and adaptable individual who is passionate about leveraging technology to create meaningful solutions, let's connect. Whether you're exploring new opportunities or just looking to share ideas, I'm always open to new connections and collaborations.

SKILLS

Here are some programming languages I use for development,
along with the technologies I am skilled in utilizing across different operating systems

PROGRAMMING LANGUAGES

  • Python
  • C
  • Java
  • JavaScript
  • MATLAB
  • SQL

SOFTWARE & TECHNOLOGIES

  • GitHub
  • SolidWorks
  • AutoCAD
  • AnsysWorkBench
  • Power BI
  • Minitab

Projects

FDM 3D Printer

Engineered a fully functional extruder and baseplate moving mechanism for an FDM 3D Printer using SolidWorks. Conducted comprehensive part selection and cost analysis. Made engineering drawings and specifications.

Gearbox Design

Designed a right-angle gearbox with a 3:1 reduction ratio in SolidWorks, applying GD&T. Implemented lean Six Sigma for rapid prototyping, reducing print time to under 6 hours and assembly to under 10 minutes.

Autonomous Driving Car with Advanced Lane and Object Detection

Utilized a CNN based on NVIDIA’s model for real-time autonomous navigation in the Udacity simulator. Integrated OpenCV and YOLOv3 for accurate lane detection and object recognition, enhancing decision-making with real-time performance metrics.

Music Genre Categorization Using Deep Learning

Performed data augmentation on GTZAN dataset, visualized the audio signals in the time domain, analyzed the frequency components of the audio signal over 0.01-second intervals by applying FFT, and generated spectrograms for the audio samples. Used feature extractor - AlexNet neural network and implemented a fully connected neural network as the classifier.

Quantitative Trading Strategy with Backtesting and Risk Management

Developed an algorithmic trading strategy in Python using 50-day and 200-day moving average crossovers combined with an Average True Range (ATR) volatility filter to generate reliable buy/sell signals. Performed comprehensive backtesting on historical stock data to optimize strategy parameters, resulting in a 50% increase in profitability compared to traditional crossover methods.

Portfolio Optimization using Modern Portfolio Theory (MPT)

Developed a Python model for optimizing a tech stock portfolio using Markowitz's Modern Portfolio Theory. Simulated 10,000 portfolios, calculated key financial metrics, and constructed the Efficient Frontier. Achieved maximum Sharpe ratio of 1.04, optimized portfolio allocation leading to a risk-adjusted return improvement. Optimized portfolio allocation: 62.47% Google, 23.75% Tesla, and 13.77% Apple.