Charles Pratt — Software Engineer

Research-oriented software engineer and CS student with end-to-end experience building and deploying AI systems — from data collection and model training through real-world execution. Experienced integrating large language models, reinforcement learning-adjacent policies, and autonomous pipelines in fast-paced research environments. Proven ability to own ambiguous projects independently and collaborate closely with researchers to turn ideas into working systems.

Education

B.S. Computer Science

University of Florida · Gainesville, FL · Aug 2023 – May 2027

Minor in Digital Arts and Sciences

Machine Learning (Grad) · Natural Language Processing · Algorithm Abstraction & Design
Autonomous Robots · Competitive Programming · Operating Systems · Info & Database Systems

GPA 3.85

Technical Skills

Languages & Tools

Python C++ Java PyTorch ROS2 OpenCV SQL React Git Linux CI/CD

AI & Research

LLMs Reinforcement Learning Diffusion Policies Imitation Learning RLHF Model Training Real-Time Inference Experiment Design

Research Experience

AI Research Engineering Intern

May – Aug 2025

Institute for Human and Machine Cognition · Pensacola, FL

Owned end-to-end development of an autonomous AI pipeline for the Unitree H1-2 humanoid — from VR-based teleoperation data collection through diffusion policy training and live deployment on physical hardware.
Integrated LLM-based planning to translate natural language task commands into structured execution goals, bridging high-level reasoning with low-level robot control.
Bridged Python-based AI frameworks (Lerobot) with Java control systems via ROS2, enabling cross-language research workflows and rapid iteration on learned policies.

Research Engineering Intern

May – Aug 2024

Institute for Human and Machine Cognition · Pensacola, FL

Built sub-100ms motion capture pipelines processing synchronized multi-sensor streams to drive real-time control of three robotic platforms: Psyonic Ability Hands, Franka Arms, and humanoid Nadia.
Developed evaluation and visualization tools that enabled researchers to rapidly interpret system behavior and iterate on experiments during live trials.

Leadership

Design Team Captain

Aug 2024 – Present

Gator Theme Park Engineering & Design · Gainesville, FL

Independently scoped and led development of an animatronic control system — designing architecture, coordinating cross-functional subteams, and delivering a working system end-to-end.
Drove agile workflows across mechanical, electrical, and software teams; built and shipped iteratively under tight timelines with real deployment constraints.

Projects

Ride Design Vehicle

Gator TPED

Electrical systems architecture for a fully inverted dark ride prototype where the vehicle hangs beneath the track. Developed a WiFi-based MQTT communication layer using Mosquitto to coordinate ride vehicles, track systems, and the central server — including servo-driven switch tracks, motorized drop-track actuation, and IR beam-break sensors for sequencing.

Learning-Based Robotic Manipulation

Autonomous Robotics Course

Modular autonomous agent pipeline with an LLM planning node that parses natural language into structured goals, a visuomotor policy for physical execution, and ROS2-based infrastructure. Ran full experimental cycle: teleoperation data collection → imitation learning → policy evaluation → hardware deployment.

Animatronic Gator Head

Gator TPED

Cross-platform control pipeline (Unity → Python → embedded) with FFT-based audio-reactive behaviors and CSV-driven sim-to-real animation deployment.

Movie Graph Visualizer

Data Structures & Algorithms

BFS and Dijkstra pathfinding over a 45,000-node movie-actor graph with an interactive HTML/JS visualization to explore shortest-path connections between films.

Jar of Joy

Software Engineering Course

Full-stack web app with Python/MySQL backend and React/JS frontend, implementing user authentication and a dynamic database-driven experience.