Running Make It Say No: three validation layers, measured in both directions 🩺 Validate pharmacy counseling requests with three safety layers
Running Answer the Patient: what a reranker changes, measured 🩺 Answer clinical questions using a handbook with citations
Running Catch the AI Lying: an audited pass/fail harness for one small free model 🩺 Audit LLM answers against gold standards
Running Order the Lab: a tool-calling trace, and what the model said with no tools 🩺 Answer drug formulary queries using a model with tool lookup
Running Agents Castab2026 Lab03 Cost Dashboard 📊 Visualize healthcare spend concentration and key metrics