# About Thomas Carr, Ph.D.

Thomas Carr is an AI researcher and software engineer based in Charlotte, North Carolina. He earned his **Ph.D. in Computing and Information Systems** from the **University of North Carolina at Charlotte** in **May 2026**, with a dissertation titled *Preserving User Privacy on Skeleton-Based Motion Data*.

## Research Focus

Thomas's research investigates privacy preservation for skeleton-based motion data in virtual and augmented reality settings. Although skeleton data appears anonymous, it encodes personally identifiable information through anthropometric structure and motion style, enabling re-identification. His dissertation contributes a family of attack and defense models that together characterize the privacy–utility trade-off:

- **Linkage Attack Neural Network (LAN)** — a Siamese classifier that re-identifies individuals across skeleton sequences.
- **Privacy-centric Deep Motion Retargeting (PMR)** — adversarially trained retargeting that suppresses identifying features while preserving action content.
- **Explanation-based anonymization** — explainable-AI techniques that localize and mask privacy-sensitive joints.
- **DisentangledTMR** — a factorized-transformer retargeting architecture with explicit identity/action disentanglement.
- **MIRAGE** — a causal, streaming autoregressive anonymizer that operates frame-by-frame with no target skeleton required.

## Current Role

**AI & Software Engineer, Incerta Intelligence** (July 2025 – present). Works on defense contracts focused on multimodal fusion for explainable and auditable decision support. Builds AI systems for mission-critical applications with emphasis on transparency and accountability.

## Education

- **Ph.D., Computing and Information Systems** — UNC Charlotte, 2023 – May 2026. Dissertation: *Preserving User Privacy on Skeleton-Based Motion Data*.
- **M.S., Computer Science** — UNC Charlotte, 2022.
- **B.S., Computer Science** — UNC Charlotte, 2019 – 2021.

## Selected Prior Experience

- **Founding AI & Software Engineer, ACR Technologies** (Dec 2024 – Oct 2025) — built an AI-enhanced intraoperative neural-monitoring communications platform.
- **Graduate Research Assistant, UNC Charlotte** (Jan 2023 – 2025) — published at leading AI/CV venues on privacy-preserving motion analysis and bias/fairness in ML.
- **Full-Stack Developer / Systems Engineer, MDcentric Technology** (May 2021 – Dec 2022) — built a real-time asset-tracking system for 30,000+ devices.
- **Founder / Lead Developer, ViBot** (Mar 2020 – Feb 2024) — a Discord platform serving 125,000 daily users, with a 98%-accuracy ML duplicate-account detector.

## Expertise

Privacy-preserving machine learning, skeleton-based motion data, motion retargeting, virtual-reality privacy, ethical machine learning, explainable AI, deep learning, computer vision, multimodal fusion, full-stack development.

## Identifiers

- ORCID: [0009-0006-6039-0209](https://orcid.org/0009-0006-6039-0209)
- Google Scholar: [scholar.google.com/citations?user=a1uc2zEAAAAJ](https://scholar.google.com/citations?hl=en&user=a1uc2zEAAAAJ)
- GitHub: [github.com/thomasc33](https://github.com/thomasc33)
- LinkedIn: [linkedin.com/in/thomasc33](https://www.linkedin.com/in/thomasc33/)
- Email: thomas@thomasc.tech