The Latest from Latent AI

Register

Join our Webinar on October 19 with Latent AI’s Director of AI, Jan Ernst !

With the LEIP SDK, developers can more easily and quickly build machine learning models by leveraging ML Application Frameworks to add AI to their applications. ML Application Frameworks are pre-validated solutions configured for common edge architectures, greatly speeding model training, prototyping, and delivery.

Learn more about how to stop building models from scratch and jump-start your edge AI during our webinar, “How to Simplify your Model Design, Exploration, and Delivery with ML Application Frameworks”.

Click here for registration details.

Events

September 6 (National Harbor, MD): The Defense Strategies Institute AI for Defense Summit brought together leaders from OSD CDAO, US Army Project Linchpin, US Navy Project Overmatch, and industry to share their perspectives. Panelists discussed the need for AI on the tactical edge where response times are critical.

Tactical Edge AI for JADC2 Panel:

Moderator: Bharat Patel , US Army PEO IEWS, Product Lead, Project Linchpin

Panelists:

• Col. Matt Strohmeyer "Nomad", Director JADC2 GIDE, DoD Chief Digital and Artificial Intelligence Office (CDAO) (CDAO)
• Charles Yetman Senior Scientific Technical Manager (SSTM) for Data Science at NIWC Pacific
• Eric Syphard , VP of AI Engineering, Booz Allen Hamilton
• Dr. Sek Chai , CTO & Co-Founder of Latent AI Watch the video below to hear what lessons corporate America and the military can learn from each other on successfully leveraging edge AI.

Thank you to Defense Strategies Institute for organizing the summit and our partner Booz Allen Hamilton for co-sponsoring this incredible panel.

Publications

Latent Runtime Engines (LREs)

LREs are easy to deploy executable runtime environments that combine model optimization with hardware performance to solve specific problems like aerial detection or meet your unique requirements.

Features:

• Modular Open System Architecture (MOSA) enabling plug and play solution with no additional tooling required ?
• Combine model optimization and compression with your hardware performance target requirements for more mission per device (extend flight time, add new capabilities, etc.)
• Build in industry leading model security to prevent model inversion, tampering, etc.

Read more here.

Latent AI Tech Highlights

Latent AI ML Engineer Sarita Hedaya sits down with intern Amogh Joshi who recently developed a fully working prototype of a deep learning neural network watermarking model. In this video, Amogh shares his insights on developing a watermarking technique and integrating it into the Latent AI Application Framework (AF) training process. Watch the video to learn more about this exciting project and the collaboration between Sarita and Amogh at Latent AI:

Upcoming Events

AUSA, October 9-11 (Washington, DC): Find us at two booths at the Army’s most powerful event this year. Latent AI is integrating with partner hardware from One Stop Systems (OSS) and St Engineering to produce optimized and secured models more quickly and reliably.

RDAX ‘23 Dragon Spear, October 22-26 (Fort Story, VA): Latent AI has been selected to showcase in RDAX’23 (Dragon Spear). We are excited to demonstrate how we enable speed to mission to enhance SOF capability needs.