CamdenMann

Built an AI agent to autonomously play the Chrome Dino Game using Neuroevolution of Augmenting Topologies (NEAT). The model evolves neural network structures and weights over generations, learning optimal jump and timing strategies without supervised data. Achieved consistent high scores through emergent behavior and fitness-based selection.

Full-stack course scheduling tool actively used by Cal Poly students. Built a Scrapy web scraper to pull live course data, professor ratings, and seat availability into PostgreSQL. Implements conflict-free schedule generation with time constraint and preference filters.

Built in under 32 hours, Constryke is an AI-driven fraud detection system designed to prevent real estate wire transfer scams. The platform parses escrow documents and email communications, extracts structured transaction data, and analyzes email metadata to detect Business Email Compromise patterns such as last-minute instruction changes, pressure language, routing mismatches, and domain impersonation.

Full-stack restaurant ordering platform with user auth, cart, and Stripe checkout. Express + PostgreSQL REST API handles menu, pricing, and inventory. Currently architecting a real-time sync layer to pull live data from POS systems.

Led a team building an AI-powered algebra tutoring platform. Architected the full-stack system (Express + MongoDB Atlas) and integrated an LLM-backed chatbot that delivers real-time hints and step-by-step error explanations.

Building simulation infrastructure for a student-designed lower-limb exoskeleton to aid individuals with mobility challenges. Modeling joint kinematics and validating motor control algorithms in ROS2/Gazebo before hardware deployment.