Hi, I'm Joshua Mason.
I'm a full‑time Applied Scientist at Thomson Reuters, where I build agentic AI systems for tax, audit, and accounting. I focus on LLM assistant design, retrieval workflows, and robust evaluation so that experts can trust and adopt these tools.
I completed my MCS at NC State in Dec 2024 (Park Scholar), after prior internships in computer vision (BAE Systems) and NLP data/annotation (LAS). I enjoy practical ML: Python, PyTorch, scikit‑learn, AWS SageMaker, and Spark; with an emphasis on clear objectives and measurable outcomes. I also led the cybersecurity club, studied Mandarin (Advanced‑Mid OPI), and tinker with RL and search projects.

Experience
-
•
Thomson Reuters
Applied Scientist
– Present Eagan, MN- Build agentic AI systems for tax, audit, and accounting domains.
- Design LLM orchestration, retrieval workflows, and robust evaluation/guardrails.
- Ship practitioner-focused assistants grounded in NLP/RAG best practices.
-
•
Thomson Reuters
Applied Science (NLP/ML) Intern
– Eagan, MN- Designed zero-shot NLP classification for a RAG pipeline using LangChain + AWS Bedrock.
- Built dataset and stress-tests for dynamic context injection; evaluated RAG failure modes.
- Proposed debugging experiments and metrics to improve retrieval and answer quality.
-
•
BAE Systems, Inc.
Data Science (R&D) Intern
– Durham, NC- Researched cloud segmentation for multi-band live streams; OpenCV, scikit-learn, PyTorch.
- Built dataset and U-net variants; achieved >75% accuracy at 15 FPS across visual/IR.
-
•
Laboratory for Analytic Sciences
Data Science Intern
– Raleigh, NC- Scraped Mandarin-language news for NLP event extraction; annotated and ran inference.
- Applied Mandarin skills for dataset quality and evaluation.
-
•
Punch Cyber Analytics Group
Cybersecurity Intern
– Reston, VA- Engineered weak-labeling for 1TB+ Zeek network logs with Apache Spark.
- Performed OSINT investigations with Maltego on APT threats.
-
•
People for PSEO
Policy Research Analyst; Chair, Student Committee
– Minneapolis, MN- Researched education data and built visualizations to support policy advocacy.
Education
-
•
North Carolina State University
MCS, Computer Science (GPA 3.93)
– Raleigh, NC- Park Scholar; AI, Data Mining, Convex Optimization, Databases, Software Engineering, Cloud Architecture.
-
•
North Carolina State University
B.S. Computer Science; Minor in Economics (GPA 4.00)
– Raleigh, NC- Summa Cum Laude; Park Scholar; Critical Language Scholar; President, Cybersecurity Club.