Bio

You can find my one-page resume on the right.

Basics

Name Minglai Yang
Label UGRA in @IVILAB, CLULAB and ML4AI-LAB｜President@AI Club｜CS&Math@UA｜Ex-MLE Intern@Coretechs
Email mingly@arizona.edu
Phone (240) 453-1294
Url https://ymingl.com/
Summary A top-performing computer science student at the University of Arizona with extensive experience in AI, software engineering, and statistical machine learning.

Education

  • 2023.08 - Present

    Arizona, USA

    Bachelor of Science (BS)
    University of Arizona, Tucson, Arizona
    Computer Science, GPA: 4.0/4.0
    • Computer Vision
    • Deep Learning for Natural Language Processing
    • Software Development
    • Computer Organization
    • Object-Oriented Programming (OOP)

Work

  • 2024.05 - 2024.08
    Machine Learning Engineer Intern (Full Time)
    CoreTechs Consulting Inc.
    Enhanced website functionality and user engagement through AWS services and intelligent bot integrations.
    • Orchestrate end-to-end website development, including design, deployment, and management on AWS with Nginx, Docker, and services like EC2, RDS, and S3.
    • Utilize HTML, CSS, JavaScript, and PHP, alongside GitLab for version control and deployment.
    • Build intelligent bots for Discord and Slack, employing GPTs and Pinecone for natural language processing and document retrieval, integrating Slack into official websites through live chat in Javascript.
    • Create React “Live-Chat” applications to optimize workflows and enhance user experience.
    • Simulated 500 concurrent users with Selenium to rigorously stress test the web application's performance.
  • 2024.03 - Present
    AI Engineer
    The University of Arizona, Institute for Computation and Data-Enabled Insight
    Advanced AI capabilities in GPT models, computer vision, and gaming applications, driving efficiency and performance gains.
    • Lead and optimize GPT-based models to process 40% more tokens per request, reducing operational costs by 25% while maintaining high model performance.
    • Increasing accuracy of roadway defect detection by 15% using OpenCV library and YOLOv10 model.
    • Acquire client data and establish a database on AWS RDS, improving data analysis efficiency by 20%.
    • Develop an AI for Minecraft using video training, reinforcement, and imitation learning, achieving 25% improvement in task execution speed and adaptability in dynamic environments.
  • 2023.09 - Present
    Research Assistant
    The University of Arizona, ML4AI LAB and IVILAB
    Implemented dynamic Bayesian networks and Theory of Mind concepts in AI research, enabling accurate prediction of human behaviors and enhancing real-world AI applications.
    • Develop the application of dynamic Bayesian networks and Theory of Mind concepts within the ToMCAT project, enhancing AI's ability to interpret complex human behaviors and predict their plans and goals.
    • Collaborate with Dr. Kobus Barnard, Dr. Clayton Morrison and Dr. Adarsh Pyarelal to translate advanced machine learning research into practical, deployable AI systems for real-world impact.

Skills

Statistical Machine Learning
Probabilistic Modeling
Dynamic Bayesian Networks
Pattern Recognition
Computer Vision
Artificial Intelligence - Deep Learning
Deep Learning
Deep Learning for Natural Language Processing
Deep Learning for Computer Vision
PyTorch
Convolutional Neural Networks (CNN)
Recurrent Neural Networks (RNN)
Long Short-Term Memory Networks (LSTM)
Generative Adversarial Networks (GAN)
Transformer Networks
Deep Reinforcement Learning
Software Engineering
Python
Java
Matlab
JavaScript
Swift
PHP
HTML/CSS
MySQL
C
Cloud Services
AWS EC2
AWS S3
AWS Lambda
AWS RDS
MongoDB
Azure Virtual Machines
Azure Blob Storage
Azure Functions
Azure SQL Database
Data Visualization - Graph Drawing
Graph Convolutional Neural Network (GCN)
Graph Drawing Algorithms
Force-Directed Layout
Hierarchical Layout
Graph Visualization Tools
Interactive Graph Visualization
Network Analysis
Graph Metrics and Analysis

Projects

  • 2023.09 - Present
    ToMCAT Project ($7.5M DARPA funded), University of Arizona, ML4AI and IVILAB
    Developed AI models utilizing dynamic Bayesian networks and Theory of Mind concepts. Integrated physiological data from fNIRS and EEG, along with facial expressions, to enhance predictions of individual behaviors and team dynamics in search and rescue missions, increasing decision-making accuracy by 30%.
    • Cross-validation and correlation analysis
    • Pupil data analysis with cognitive science principles
  • 2024.05 - 2024.06
    Darts Game Website Deployment with AWS and Docker
    Designed and deployed a darts game website using Nginx and Docker. Hosted on AWS with EC2 for server management, RDS for database handling, and S3 for data storage. Managed versioning and automated deployment via GitLab, boosting workflow efficiency and reliability.
    • EC2 server management
    • RDS database handling
    • S3 data storage
  • 2024.04 - 2024.05
    Bot Development for Slack and Discord
    Leveraged RAG (Retrieval-Augmented Generation) for advanced natural language processing. Employed prompting and fine-tuning methods to customize responses and enhance performance. Integrated Pinecone for efficient document retrieval, solving hallucinations in communication channels.
    • Advanced natural language processing
    • Custom response fine-tuning
    • Efficient document retrieval
  • 2024.04 - Present
    Google Calendar Integration with ChatGPT Discord Chat Bot via Node.js
    Deployed the chatbot on Amazon EC2 for continuous operation and accessibility. Integrated the Google Cloud API using Nginx for users to schedule and manage events through Discord. Implemented and tested advanced error handling and logging mechanisms to ensure robust operation.
    • Continuous operation on Amazon EC2
    • Google Cloud API integration
    • Advanced error handling and logging

Awards

Languages

English
Bilingual Proficiency
Chinese & Shanghainese
Bilingual Proficiency

Interests

Natural Language Processing
Language Models
Text Understanding
Generation
Cognitive Modeling
Theory of Mind
Human Behavior
Belief Modeling
Machine Learning
Generalization
Uncertainty
Reinforcement Learning
LLM Reasoning & Alignment
Faithful Reasoning
Step-wise Inference
Human Alignment

References

Professor Liangming Pan (liangmingpan@arizona.edu)
Assistant Professor, School of Information, University of Arizona. Supervised my research on reasoning in large language models.
Professor Mihai Surdeanu (msurdeanu@arizona.edu)
Professor, Department of Computer Science, University of Arizona. Advised me on efficient architectures for LLM inference.
Professor Kobus Barnard (kobus@arizona.edu)
Professor, Department of Computer Science, University of Arizona. Mentored me for several years on probabilistic graphical models and cognitive science.
Professor Reyan Ahmed (abureyanahmed@arizona.edu)
Assistant Professor, Department of Computer Science, University of Arizona. Prof. Ahmed was my mentor during my research on a graph drawing project.
James Bloomer (jbloomer@coretechs.com)
Partner and Vice President of Coretechs. I worked closely with James during my internship at Coretechs, where he oversaw my contributions to projects.