GARDE-Chat: A Scalable, Open-Source Platform for Building and Deploying Health Chatbots

Presentation Type: Systems Demonstration
Not recorded for AMIA Now

Presentation Time: 03:30 PM - 03:55 PM

Primary Track: Clinical Research Informatics


GARDE-Chat is an open-source platform enabling scalable, chatbot-based digital health interventions. It supports the development of both rule-based and LLM-driven chatbots, multilingual delivery, integration with external systems such as electronic health records (EHR) and REDCap. GARDE-Chat has powered interventions in large clinical studies across genetic testing, COVID-19 testing, and tobacco cessation. This demonstration showcases GARDE-Chat’s architecture, authoring workflow, and live examples illustrating its capacity to enhance patient engagement and digital health research.

Speaker(s):
Guilherme Del Fiol, MD, PhD
University of Utah

Author(s):
Guilherme Del Fiol, MD, PhD - University of Utah;

---

AID-AD with AD Commons and ChatAD: Multimodal AI Collaboratory for Alzheimer’s Disease & Related Dementias

Presentation Type: Systems Demonstration
Not recorded for AMIA Now

Presentation Time: 03:55 PM - 04:20 PM

Primary Track: Data Science/Artificial Intelligence


The AID-AD Collaboratory is a multimodal AI initiative accelerating discovery and translation in Alzheimer’s disease and related dementias. Its core platform, AD Commons, integrates imaging, omics, and clinical data with pretrained models, standardized pipelines, and FAIR-compliant metadata, while ChatAD enables natural-language dataset discovery and study design. Our systems demonstration will show dataset search, model benchmarking, and results visualization. AID-AD advances reproducibility, interoperability, and equitable access for dementia research.

Speaker(s):
Pedro Pinheiro-Chagas, PhD
University of California, San Francisco (UCSF)

Tomiko Oskotsky, MD
UCSF

Alice Tang, PhD
UCSF

Marina Sirota, PhD
University of California, San Francisco

Author(s):
Tomiko Oskotsky, MD - UCSF; Alice Tang, PhD - UCSF; Marina Sirota, PhD - University of California, San Francisco; Pedro Pinheiro-Chagas, PhD - University of California, San Francisco (UCSF); Michael Weiner, MD - VA Medical Center, San Francisco, CA;

---

Data We Need To Chat: AI-Powered Tools for Biomedical Data Discovery and Agent-Based Research

Presentation Type: Systems Demonstration
Not recorded for AMIA Now

Presentation Time: 04:20 PM - 04:45 PM

Primary Track: Data Science/Artificial Intelligence


Synapse Chat and Synapse MCP Connector are AI-powered tools enabling natural language interaction with Synapse.org's 3.7PB+ biomedical data repository. Synapse Chat uses Retrieval-Augmented Generation for conversational queries. At the same time, the MCP Connector, featured in Anthropic's "Claude for Life Sciences" initiative, enables AI agents to access entities, annotations, and provenance programmatically. Both implement tiered authorisation, ensuring HIPAA/GDPR compliance while enabling "metadata spelunking" without accessing sensitive data.

Speaker(s):
Susheel Varma, PhD MBA FBCS
Sage Bionetworks

Author(s):
Susheel Varma, PhD MBA FBCS - Sage Bionetworks; Jineta Banerjee, PhD - Sage Bionetworks; Robert Allaway, PhD - Sage Bionetworks; John Hill, BS - Sage Bionetworks; Jay Hodgson, PhD - Sage Bionetworks; Alberto Pepe, PhD - Sage Bionetworks; Christine Suver - Sage Bionetworks; Luca Foschini, PhD - Sage Bionetworks;

---