Blake Pennel
Computer Science and Mathematics at the University of Kansas. I build agentic LLM systems and am heading towards interpretability research.
- [email protected]
- github.com/Blake192
- linkedin.com/in/blakepennel
- Lawrence, Kansas
- blakepennel.com
Education
University of Kansas · B.S. Computer Science, B.S. Mathematics
Aug 2024 – May 2028SELF Fellowship · Honors · GPA 3.9
- Study abroad at Korea University, with coursework in AI and deep learning.
Experience
AI & Cybersecurity Intern · NASA Jet Propulsion Laboratory
Jun – Aug 2026 · Pasadena, CA- Designed and developed an agentic AI application for NASA’s Deep Space Network that analyzes cybersecurity vulnerabilities and generates actionable remediation recommendations (write-up).
- Researched and evaluated tradeoffs in model integration, application architecture, and cybersecurity requirements, translating findings into system design and implementation decisions.
Machine Learning Research & Development Intern · MITRE
May – Aug 2025 · McLean, VA- Researched, prototyped, and delivered machine learning solutions, evaluating hardware, software, framework, and model tradeoffs to produce a sponsor-facing deliverable and an end-to-end demo.
- Developed a repeatable process for integrating machine learning into existing projects, and collaborated with multiple teams to add AI capabilities to their systems.
Research & Development Intern · Cornerstone Integration
Jun 2023 – Jan 2025 · Leavenworth, KS- Designed, prototyped, and demonstrated integrated hardware and software systems for research, evaluation, and partner-facing efforts.
- Researched and implemented AI, computer vision, embedded and IoT, cloud, networking, and security technologies with vendor engineering teams to guide technical integration decisions.
Projects
Location memory for AI assistants
Personal · 2026Self-hosted MCP servers over my own Google Maps Timeline; found and fixed an undocumented multi-part backup format, recovering all 8,551 segments (write-up).
Technical report
Blake Pennel. “Automating Cybersecurity Recommendations for Remediation of Vulnerabilities within NASA’s Deep Space Network: An Agentic Approach.” NASA Jet Propulsion Laboratory, California Institute of Technology, 2026. Technical report. PDF
Recognition
- Two-time CyberPatriot National Finalist, ranking in the top 0.1% nationally among 5,264 teams.
- Programmed a FIRST Tech Challenge robot for manual and autonomous operation with computer vision.
Skills
- ML & deep learning
- PyTorch, computer vision, CNNs, Transformers, LLMs, fine-tuning, RAG, sequence models, edge AI, LiteRT, ONNX Runtime, LightGBM
- Agents & LLM systems
- Model Context Protocol, Pydantic AI, structured output, tool use
- Programming
- Python, C, Java, TypeScript, Bash, PowerShell
- Systems
- Linux, Docker, Git, virtualization, Nginx, FastAPI, React, AWS, Azure