Last year's AI <> Tradecraft Forum was a huge success, so we are expanding it this year!
Thought Leader Presentations (All Day):
Speakers and Presentations are Subject to Change
| Dr. Pat Biltgen | Director of Analytics, Perspecta and Author of Activity Based Intelligence |
| David J. Gauthier, Director, Commercial and Business Operations Group, NGA Source | “Reaching all the data for GEOINT” |
Walter Paz, Department of the Army, G2 | “OSINT challenges, successes and the road ahead” |
Hannah MacKenzie-Margulies, Product Manager, Basis Technology, Rosette Text Analytics | “Multilingual Analytics for Open Source and Dirty Data” |
| Andrew E. Newton, J2 Technical Director, Cyber National Mission Force, USCYBERCOM | “To index or not to index? That is the data lake question” |
| Robert Shelton, Jr., CTO, Microsoft Federal | “How AI and continuous monitoring affect force protection” |
Dr. Ian Soboroff, Group Leader, Retrieval Group, Information Access Division | “TREC: measuring the effectiveness of information analytics” |
Bryan Stoker, Tech Director, NSA Operations Center (NSOC) | TBD |
| CW3 Nathan McKeldin, Senior Trainer/Curriculum Developer, Army OSINT Office | “Artificial Intelligence and the Intelligence of the Future: How Ai Can Improve the Existing Intelligence Cycle and What It Means for Analysts” |
Hands-On Workshops (Afternoon):
Open Source Web Research: Hands-On PAI Tradecraft and Tools
- Focusing on the countries of India and Russia, Diffeo’s analyst team will teach the latest PAI exploitation techniques to research foreign government interest in critical DoD supply chains and political leadership goals Tools include: managed attribution with Authentic8/Toolbox, Recorded Future, Basis multi-lingual analytics, and more.
Machine Learning Tutorial: Python Vector Embeddings for Entity-Entity Relations
- Diffeo’s machine learning team will provide a hands-on tutorial based in an iPython notebook adapted from our colleague Sameer Singh’s tutorial on vector embeddings for entity-to-entity relations. This clever use of matrix factorization combines unstructured documents with relations from a structured knowledge base to create dense vectors that encode the semantic meaning of relationships.
