F2: ML 102 Technology Developments in AI/ML: From Prompts to Agents

Time:
Monday, June 1, 10:45 a.m. - 12:15 p.m.
Location: Meeting Room 1-3

Building on ML 101 foundations, this tutorial surveys the current AI/ML landscape, distinguishing genuine value from hype. As AI integrates invisibly into daily workflows, understanding practical utility versus high-level marketing becomes critical. We explore effective tools and techniques while navigating significant pitfalls in output quality and data security.

The curriculum challenges industry buzzwords, specifically questioning "Prompt Engineering" as a lasting discipline. While making this case, we provide deep dives into effective prompting techniques and parameters that influence model behavior.

The central focus is on LLM evolution from reactive generators to Agentic AI. Participants explore how models approach complex problems and transition from using AI as a search engine to a proactive partner.

Mechanism transparency grounds every discussion in underlying mechanics. By lifting the veil of mystery surrounding these technologies, this course returns agency to users, transforming them from passive consumers into informed, strategic operators of AI systems.

Renee Yazdi is the deputy lead engineer for Canyon, where she primarily supports Air Force Research Laboratory (AFRL) PNT projects, SWAC, SDUE, and Internal Research and Development (IR&D) efforts. With 30 years of experience, she has extensive expertise in the space domain, contributing to a diverse range of systems engineering and technology development initiatives. Her technical expertise includes remote sensing, communications, missile systems, and GPS/GNSS technologies. Ms. Yazdi holds a MS in EE Cal Poly, as well as a certificate in Artificial Intelligence and Machine Learning from Caltech.

Brian Zufelt is the deputy program manager for PNT programs at Canyon with over 18 years of experience in embedded systems, computer vision, and machine learning. He holds an MS in Computer Engineering from UNM and previously led ML platform development for GPS threat detection systems at COSMIAC, supporting AFRL, NASA, and DoD partners.