Dustin J. Ober
Data Scientist & ML Engineer
Chantilly, VA
Data Scientist specializing in NLP, Machine Learning, and Production ML Systems. 30+ models deployed for defense and intelligence applications, bridging educational strategy with AI implementation.
Core Specializations
Current Position
ML Engineer / Data Scientist
Leidos - Chantilly, VA
- 97% Automation: Built data pipelines reducing report generation from 1 week to 30 minutes
- Custom NER Models: Designed Named Entity Recognition models for specialized intelligence datasets
- Production ML: Trained and deployed 30+ ML models using TensorFlow and DSPy
- Accessibility Leadership: Created Section 508/WCAG database adopted organization-wide
- ML Infrastructure: Built secure FastAPI REST services with JWT and Dockerized applications
Professional Credentials
Project Management Professional (PMP)
PMI
2024
TensorFlow Developer Certificate
2023
AWS Cloud Practitioner
Amazon
2023
IBM Data Science Professional
IBM
2022
Google Data Analytics
2022
Published Research
Author of The Sovereign AI Handbook series - technical whitepapers on deploying AI systems in air-gapped, classified, and compliance-heavy environments.
Sovereign AI Infrastructure
Hardware & Architecture for Disconnected Environments
The Disconnected Pipeline
Solving Dependency Management in Secure Facilities
Private Knowledge Retrieval
Architecting Local RAG Systems
Verifiable Intelligence
DSPy, Governance, and Hallucination Control
Security Clearance
Top Secret Clearance with SCI and Polygraph
Trusted professional working on sensitive defense and intelligence projects requiring the highest levels of trust and expertise.
Frequently Asked Questions
How do you implement AI in classified environments?
I specialize in deploying AI systems within air-gapped, classified networks using sovereign infrastructure patterns. This includes local model hosting, containerized deployments (Docker/Apptainer), and secure dependency management without internet access.
What is sovereign AI infrastructure?
Sovereign AI refers to AI systems that operate entirely within controlled boundaries—no cloud dependencies, no data egress. It's essential for defense, intelligence, and regulated industries requiring 100% data sovereignty.
Do you work with defense contractors?
Yes, I have extensive experience working with defense and intelligence organizations. I hold a Top Secret clearance with SCI and Polygraph, enabling me to support sensitive mission-critical projects.
How long does AI training curriculum development take?
Typical AI training curriculum development ranges from 4-12 weeks depending on scope. I use the ADDIE framework combined with rapid prototyping to deliver comprehensive technical training efficiently.
Let's Work Together
Available for consulting on AI implementation, instructional design, and technical training development.