My name is Jordan Hunt.
I build intelligent systems.
Graduating Computer Science senior at Tuskegee University and incoming Georgia Tech OMSCS student, focused on reinforcement learning, anomaly detection, and end-to-end ML pipelines. I build things that learn, from Q-Learning agents to AI-powered civic tech platforms, and I care about making AI accessible and useful for real people.
Where I've Studied
A foundation in theory, applied through building.
Georgia Institute of Technology
Tuskegee University
Where I've Worked
Teaching, research, and real-world problem solving.
Teaching Assistant, Programming II (C++)
- Supports 25+ students per semester in advanced programming concepts including inheritance and polymorphism.
- Provides tutoring and hands-on lab assistance to improve student comprehension and project outcomes.
Data Science Researcher (Python)
- Applied techniques from UC Berkeley coursework to analyze racial disparities across 10+ years of college education data.
Student Researcher (Python)
- Developed and optimized an AI autoencoder achieving 97%+ detection accuracy on large-scale anomaly detection datasets.
Resident Assistant
- Organized housing operations for 75+ students, planning 8+ community events.
Optimal Q-Learning Agent
A reinforcement learning agent that learned Tic-Tac-Toe from scratch through self-play and opponent-pool training.
Optimal Q-Learning Agent
Built a near-optimal Q-Learning agent for Tic-Tac-Toe via self-play and opponent-pool training (Temporal Difference + Bellman). Engineered perspective-agnostic state representation and full board symmetry to compress the state-action space by ~6-8x, allowing faster convergence with fewer unique states. Deployed as a real-time human-playable demo.
What Else I've Built
From data analysis to NLP classification.
RecipeSnap
Mobile-first web app that identifies ingredients from photos using AI vision (Qwen VL 72B via OpenRouter) and generates personalized recipes. Built an image compression pipeline with Pillow and defensive JSON parsing for unreliable LLM outputs.
Higher Education Retention Analysis
Collaborated in a 5-person research team to analyze 10 years of federal data across 6 racial demographics affecting college enrollment. Cleaned 3 federal datasets and created 6+ data visualizations.
.png)
Spam Classifier
Built a text classification pipeline using TF-IDF feature extraction with Logistic Regression and Random Forest models for spam detection. Compared model performance across precision, recall, and F1 metrics.
Competition Wins
Back-to-back 1st place finishes in 2026.
InternXL AI Innovation Challenge
Led a team to 1st place building an AI-powered civic tech mobile app targeting young voters. Integrated LLM-backed personalization across local, state, and federal levels using real-time geolocation for tailored content delivery.
HBCU Battle of the Brains
Built a Chrome extension and analytics platform within an 8-person team. The extension scores user prompts against 50+ shopping keywords to detect intent, enhancing AI responses with product cards. Engineered the data pipeline powering analytics for brands, delivering a full MVP in 24 hours.
What I Work With
Tools, languages, and concepts I use to build.
Languages
AI / Machine Learning
Tools & Frameworks
Concepts
Beyond the Code
Teaching, mentoring, and leading teams.
Chief of Staff, ACM E-Board
Leads coordination of study nights and executive board operations for the campus ACM chapter.
Instructor
As a Black Belt, mentors and instructs students of all ages, fostering discipline and leadership through martial arts training.