---
title: "How the FDA Built an AI Platform That 85% of Its Staff Now Use Daily"
description: "The FDA built ELSA, a generative AI platform for its 16,000 staff, on Halo, a governed Databricks data foundation created to address fragmented systems across eight centers. Those centers had separate chatbots and data stores; consolidating 50 to 60 sources enabled faster sharing, real-time streaming, and centralized access controls through Unity Catalog. Within roughly two months, ELSA adoption rose from less than 1% to 85%, while staff began building hundreds of agents weekly from standard operating procedures, regulatory guidance, and center-specific documents. MCP servers layered over Unity Catalog make governed data and tooling accessible beyond data scientists, and Databricks ML and NLP capabilities through MLflow extracted starting materials and product-supplier-manufacturer relationships from millions of submission pages. A reviewer can now request grounded starting-material information for a drug application in about three minutes instead of days, while the FDA adapts center-specific MCP tools and extends the model across its organization."
---

# How the FDA Built an AI Platform That 85% of Its Staff Now Use Daily

[Databricks](https://yomu.fyi/company/databricks) · Molly Just-Behr · Jul 23, 2026

**Type:** Explainer

## Summary

The FDA built ELSA, a generative AI platform for its 16,000 staff, on Halo, a governed Databricks data foundation created to address fragmented systems across eight centers. Those centers had separate chatbots and data stores; consolidating 50 to 60 sources enabled faster sharing, real-time streaming, and centralized access controls through Unity Catalog. Within roughly two months, ELSA adoption rose from less than 1% to 85%, while staff began building hundreds of agents weekly from standard operating procedures, regulatory guidance, and center-specific documents. MCP servers layered over Unity Catalog make governed data and tooling accessible beyond data scientists, and Databricks ML and NLP capabilities through MLflow extracted starting materials and product-supplier-manufacturer relationships from millions of submission pages. A reviewer can now request grounded starting-material information for a drug application in about three minutes instead of days, while the FDA adapts center-specific MCP tools and extends the model across its organization.

## Context

The FDA's eight centers had developed separate AI capabilities, data stores, and chatbots, creating duplicated costs and slowing data sharing. Sensitive trade secrets and regulatory data also required strict access controls before centers could adopt a shared platform.

## Approach / What changed

The FDA consolidated 50 to 60 data sources from all eight centers into Databricks, using Unity Catalog for governance and granular table-level access. It deployed ELSA across the agency, layered MCP servers on the governed foundation, and used Databricks ML and NLP capabilities through MLflow to extract regulatory data for grounded answers and agents.

## Takeaways

- ELSA adoption increased from less than 1% to 85% of FDA staff within roughly two months of launch.
- Staff create hundreds of agents per week by loading standard operating procedures, regulatory guidelines, and center-specific documents into ELSA workspaces.
- Extracting starting materials and product-supplier-manufacturer relationships from millions of pages reduced one drug-review search task from days to about three minutes.

**Tags:** [AI Agents](https://yomu.fyi/topic/ai-agents), [AI Governance](https://yomu.fyi/topic/ai-governance), [Databricks](https://yomu.fyi/topic/databricks), [MCP](https://yomu.fyi/topic/mcp)

- Source: [Databricks](https://www.databricks.com/blog/how-fda-built-ai-platform-85-its-staff-now-use-daily)
- Source URL: https://www.databricks.com/blog/how-fda-built-ai-platform-85-its-staff-now-use-daily
- Ingested by Yomu: 2026-08-30T16:53:30.944Z

[Read original post](https://www.databricks.com/blog/how-fda-built-ai-platform-85-its-staff-now-use-daily)