Building Agentic AI Tutors for Life Sciences — RGMBS 2026

Session 01 at the Riyadh Global Medical Biotechnology Summit 2026, presented by Dr. Omar Saleem, MD. Roughly 2h 45m, fully hands-on, pods of 4–6, no coding required. Turn any expertise into an adaptive learning system.

What you build is a manifest

The whole hands-on reduces to editing one declarative file — domain, model, language, persona, knowledge pack, tools, assessor, governance and evals. You start from a running tutor and edit it; nobody faces a blank file.

Run of show

• Framing & anatomy — the facilitator builds one tutor live, end to end, as the anchor.
• Build Block 1 — make it teach, and stay on-source. Pods fork the running tutor into their own domain and watch it refuse instead of hallucinate.
• Build Block 2 — make it adaptive, and prove it's safe. Add the assessor, per-learner memory, budgets and the eval gate.
• Showcase — one harness, many tutors. Swap the brand, the language and the domain live. Config, not fork.
• Close — dynamic workflows, then the