Enabling Defense Missions with Local LLMs - Part 2

Explore the challenges and solutions for deploying Large Language Models in highly sensitive and regulated environments through this conference talk. Learn about LeapfrogAI, an open-source toolkit designed to meet rigorous requirements for running and observing models in edge and air-gapped settings. Discover the complexities of obtaining Authorization to Operate for both software and models in defense missions. Gain insights into enabling safe and successful use of Generative AI in defense applications from Gerred Dillon, an experienced open source developer and Unicorn Engineer at Defense Unicorns. Understand the unique considerations for fine-tuning, embedding sensitive data, and creating small models for specialized environments. Delve into the speaker's extensive background in AI/ML, edge compute, distributed systems, and infrastructure engineering to grasp the intricacies of implementing generative AI in local, sensitive, and airgapped environments.