TEYJ KRISHNAN

University of Michigan

B.S. Computer Science

> Machine Learning | Full-Stack Dev | Data Analysis
> Expected Graduation: May 2028
> Status: ONLINE_

About Me

University of Michigan

B.S. Computer Science
Minor in Mathematics

Duration: Aug 2025 - May 2028

Location: Ann Arbor, MI


Coursework:

Discrete Mathematics, Data Structures + Algorithms, Linear Algebra


Activities:

Kappa Theta Pi, Michigan Venture Club, Product Motion, Quant Finance Club

I'm a Computer Science student at the University of Michigan, expected to graduate in May 2028. I have a curiosity for machine learning, data analysis, and building impactful applications.

I've worked on projects in medical imaging research, computational neuroscience, and quantum computing simulations. My experience spans from developing CNN-based pipelines for low-radiation X-ray imaging to building EEG signal classifiers for motor restoration research.

Beyond academics, I'm actively involved in Kappa Theta Pi (Tech Fraternity), Michigan Venture Club, Product Motion, and Quant Finance Club, where I collaborate with like-minded individuals to explore the intersection of technology and innovation.

>>> Passionate about leveraging AI and machine learning to solve real-world problems!

Technical Skills

[ Languages ]

Python Java JavaScript C++ HTML/CSS

[ Libraries & Frameworks ]

FastAPI PyTorch TensorFlow NumPy Pandas Scikit-Learn Matplotlib React Native

[ Focus Areas ]

Machine Learning Full-Stack Development Data Analysis

Experience & Projects

Project description

Quantum Diffusion Simulator

November 2025
  • Built a quantum walk simulator using Qiskit and NumPy to model network diffusion dynamics, achieving stable simulations on up to 16-node graphs over 20 steps for direct comparison with classical random walks
  • Engineered a modular Python pipeline with NetworkX, Matplotlib, and Seaborn, enabling 3× faster data visualization and metric computation through vectorized operations and reusable graph modules
  • Analyzed diffusion behavior using entropy (2.94 vs 2.57 bits), coverage (50% vs 43.6%), and total variation distance (0.33), revealing quantifiable interference-driven divergence between quantum and classical diffusion
View on GitHub →
Project description

Medical Imaging Research

Research Assistant
Remote
Aug 2024 - Feb 2025
  • Designed a CNN-based computer vision pipeline to simulate low-radiation chest X-ray images, enabling experimentation with reduced-dose imaging scenarios and reducing experiment time by 40%
  • Applied a denoising model implemented in NumPy and PyTorch to reconstruct medical image outputs, improving perceptual quality compared to raw low-radiation simulations
  • Validated model on 200-image MedMNIST dataset, attaining 26.55 dB PSNR and 0.92 SSIM
Link coming soon...
Project description

Lumbar Degeneration Classifier

Jul 2024 - Aug 2024
  • Trained and tested deep learning models (CNN, EfficientNetV2, ConvNext) on 147,000+ MRI scans representing 2,000 patients to classify severity of spinal degeneration
  • Led development of classification model in PyTorch to cover 5 degeneration levels and 3 spinal regions, achieving 0.95 micro-average AUC for multi-class, region-specific spine analysis and 87% classification accuracy
  • Built and deployed a full-stack FastAPI web application with a clinician-focused interface for real-time MRI inference and visualization, hosted on Render with integrated model selection and probability outputs
View on GitHub →
Project description

Maddux Mortgage

Software Engineering Intern
Remote
May 2024 - Dec 2024
  • Built a client-facing mobile application in React Native to deliver real-time mortgage market insights
  • Collaborated with business founder and CEO to implement features for comparing mortgage plans and calculating prices, streamlining decision-making process for non-technical users and prospective customers
  • Utilized Figma to enhance app usability and design responsive front-end components, enabling smooth navigation across different devices and screen sizes
Link coming soon...
Project description

Computational Neuroscience Research

Lead Researcher
Redmond, WA
May 2023 - Apr 2024
  • Processed 350,000+ EEG samples with PyTorch and NumPy for signal analysis and reduced preprocessing
  • Developed EEG signal classifier achieving 98.15% accuracy and visualized neural activity with Matplotlib, validating classifier predictions and demonstrating potential for motor restoration for paralysis patients
  • Published research paper in Journal of High School Science and won 1st Place Award at WA State Science + Engineering Fair
View publication →