AI That Works for You & Building a Private LLM

---

AI That Works for You: Tools, Prompts, Privacy — and How A Warship Taught Itself History

Generative AI is everywhere — but using it effectively, safely, and without surrendering your data is a skill most people are still developing. This quarter's OEB LMAG lunch brings together and local Higher Ed leader and a CIO, two seasoned practitioners for a practical, hands-on session on getting real value from AI tools. Roger Doering, Professor of Engineering at Cal State East Bay, leads the main program on leveraging generative AI effectively and safely — covering tools, approaches, prompting strategies, and critical privacy and security considerations. He's joined by Mark Rowell, CIO of the USS Hornet Museum, who adds a fascinating real-world dimension: what happens when you build your own private LLM on a historic warship's institutional knowledge. Live demos included. Bring your questions — and your appetite.

Generative AI tools have moved from novelty to necessity — but the gap between casual use and genuinely effective use is wide. For users with concerns about accuracy, privacy, security, and institutional knowledge, the public-cloud nature of most AI tools raises real questions.
This session addresses both sides of that challenge. Roger Doering brings academic and professional rigor to the question of how to get the most from generative AI — what works, what doesn't, and how to stay safe while doing it. Mark Rowell then demonstrates what the far end of the privacy spectrum looks like in practice: a fully private LLM, built on Google's NotebookLM, trained on 165 sources of institutional knowledge about one of America's most historic naval vessels.
Together, they cover the landscape from everyday AI productivity to purpose-built institutional intelligence — with plenty of time for hands-on exploration, demos, and the kind of frank conversation that only happens over a good lunch.

WHAT YOU'LL COVER

•      Generative AI tools landscape — what's worth your time in 2026 and why

•      Prompting that actually works — practical approaches and live examples

•      AI privacy and security — what you're sharing, with whom, and how to protect yourself

•      Hands-on demos — try it yourself, not just watch it, so bring a laptop

•      The private LLM case study — how the USS Hornet built a closed, curated AI from 165 institutional sources

•      Real-world use cases from the Hornet's deployment — docents, educators, researchers, and visitors

•      Open discussion — bring your own AI questions, frustrations, and experiments

 

---

---

---

---

Order & Serve Lunch: 11-12
Meeting Topic and discussion: 12-2

---