Session
Scaling Human Connection with Real-Time AI Conversation Monitoring
Overview
| Experience | In Person |
|---|---|
| Track | Artificial Intelligence & Agents |
| Industry | Healthcare & Life Sciences, Public Sector |
| Technologies | Databricks SQL, Unity Catalog, Lakebase |
| Skill Level | Intermediate |
With over 300 million messages exchanged since 2013 and more than a decade of expertise, Crisis Text Line is now leading the implementation of LLMs in the mental health nonprofit space. We will discuss a set of conversation monitoring capabilities built on the Databricks ecosystem that provide unprecedented observability into live, high-stakes conversations. These models surface risk signals, quality metrics, and contextual insights. We will demo how these capabilities enable co-pilot-like support that helps crisis responders and clinical supervisors focus on what they do best—supporting people in need.
Session Speakers
Mateo Garcia Pepin
/Lead Data Scientist
Crisis Text Line