h2oGPTe Agent Demo: Fibromyalgia Research Report
- Agentic AI
- Enterprise h2oGPTe
This lesson walks through a medical research use case for h2oGPTe agents, focused on compiling evidence for a fibromyalgia diagnosis report. It covers setting the agent's persona and instructions across four phases: a scholarly literature search limited to recent reviews and treatment studies, an evidence analysis step that weighs methodological quality, a synthesis step using stronger reasoning models, and a final report with an executive summary and clinical recommendations. The demo shows the agent's tool use and self-correction, plus the finished report, and notes the output can be refined further with follow-up prompts.
Key takeaways
- How to structure an h2oGPTe agent prompt across phases: literature search, evidence analysis, synthesis, and reporting
- Why limiting a literature search to studies from the past three years keeps a research report current
- How the agent's scholar paper search tool differs from a general internet search for gathering evidence
- The option to keep prompting the agent after a report is generated to add more sourcing detail
Start owning your AI