Aryan Patel

Graduate Student in Artificial Intelligence · Northeastern University

I work on deep learning, computer vision, and generative models — then write about how and why they work. My focus is on building scalable solutions for real-world applications.

email cv linkedin github leetcode
Aryan Patel

About

I am a graduate student in Artificial Intelligence at Northeastern University, Boston, with a concentration in Machine Learning. My research interests lie at the intersection of deep learning, computer vision, and natural language processing, with a particular focus on developing scalable solutions for real-world applications.

I earned my Bachelor's in Computer Science and Engineering with a specialization in AI and Machine Learning from Quantum University, where I built a strong foundation in ML theory and practical implementation. My experience spans image classification, neural network architectures, and production-ready ML pipelines.

When I'm not working on models, you'll find me listening to music, exploring new ideas, or catching a 49ers/SF Giants game.

Research Interests

Deep Learning & Neural Networks Design and optimization of CNNs, RNNs, and transformer-based architectures
Computer Vision & Image Processing Object detection, segmentation, and scene understanding
Autonomous Systems AI-driven perception, control, and sensor fusion for autonomous vehicles and robotics
MLOps & ML Pipelines Scalable model deployment, workflow automation, and pipeline development using TFX
Natural Language Processing Generative models and representation learning for language understanding
Scalable & Responsible AI Efficient ML pipelines, interpretability, and ethical model deployment

Publications

Lung Damage Detection Using Various Deep Learning Algorithms

Aryan Patel, Rohit Chauhan, Suryajeet Gupta
pdf arxiv

Projects

Lung Damage Detection System

Designed and trained a CNN for multi-class classification of lung diseases, including pneumonia and lung cancer. Integrated with a real-time ML pipeline using TensorFlow Extended (TFX) for automated preprocessing and scalable training. Potential to reduce manual diagnosis time by 30%.

TensorFlow · TFX · CNN · Seaborn

Multiclass Classification Using Neural Networks

Developed and optimized a neural network on 50,000+ multi-class images, achieving 92% test accuracy through hyperparameter tuning and architecture experimentation. 15% improvement over baseline models.

TensorFlow · Scikit-learn · Matplotlib · Google Colab

Customer Feedback Classification System

Deployed a Naive Bayes classifier to categorize 10,000+ customer feedback samples at 91% accuracy. Enhanced performance by 18% through feature engineering and iterative tuning. Precision 89%, recall 87%, F1-score 88%.

Python · Scikit-learn · NLP · Feature Engineering

California Housing Price Prediction

Built and validated regression models to predict California housing prices. In-depth EDA uncovering correlations across 10+ demographic and economic indicators. Compared linear, ridge, and lasso approaches.

Python · Pandas · Scikit-learn · Regression Analysis

Education

Masters of Science in Artificial Intelligence

Northeastern University
Concentration: Machine Learning

Bachelors of Technology in Computer Science & Engineering

Quantum University
Specialization in Artificial Intelligence and Machine Learning

Experience

AI and Machine Learning Intern

YBI Foundation
November 2023 – October 2024
  • Deployed a Naive Bayes classifier on 10,000+ customer feedback samples, extracting actionable insights from unstructured data
  • Enhanced accuracy by 18% through systematic preprocessing, feature engineering, and hyperparameter tuning
  • Evaluated with accuracy (91%), precision (89%), recall (87%), F1-score (88%), and confusion matrix analysis
  • Documented methodology and results in structured reports for knowledge transfer

Machine Learning Intern

Prodigy InfoTech
July 2023 – August 2023
  • Built regression models to predict California housing prices with strong predictive accuracy
  • Conducted EDA uncovering correlations across 10+ demographic and economic indicators
  • Streamlined preprocessing pipeline, reducing preparation time by 25%
  • Collaborated with 3-member team to compare regression approaches and select best model

Technical Skills

Programming & Tools: Python, Rust, Git, Power BI, MS Office
Frameworks: TensorFlow, PyTorch, Scikit-learn, Keras, XGBoost, Pandas, NumPy, Matplotlib, Plotly, MLflow, Seaborn, TensorBoard, Flask, FastAPI
ML Techniques: Regression, Classification, Feature Engineering, Clustering (K-Means, DBSCAN, Hierarchical), Model Evaluation
Advanced AI: Deep Learning, Neural Networks, Unsupervised Learning, Reinforcement Learning, Computer Vision, NLP
Statistics: Probability Theory, Hypothesis Testing, Statistical Inference, Time Series Analysis
Data: Data Preprocessing, Visualization, Data Structures, Algorithms

Certifications

Google Cloud Career Launchpad Data Analyst
Neural Networks and Deep Learning
Machine Learning with Python (Udemy)
Deep Learning with TensorFlow

Writing

AI Weather Forecasting's Critical Blind Spot

It can predict tomorrow's weather perfectly but fails when we need it most.

Read on LinkedIn →

When AI Meets Antarctic Exploration: A Breakthrough in Marine Biology

Scientists developed an AI system that analyzes seafloor photographs from 8 hours down to seconds — a 1000x+ speedup.

Read on LinkedIn →

Amazon's Project Rainier — The $11B Data Center Revolution

One of the boldest bets in the AI infrastructure race. What makes this truly groundbreaking?

Read on LinkedIn →