Solution
Boston Children's built what it calls an enterprise AI layer around a secure internal ChatGPT environment that spans research, clinical, and administrative teams. The hospital created a shared foundation where teams work with internal data, synthesize medical literature, and launch new capabilities under shared governance and monitoring. On the operational side, AI manages invoice intake, routing, and responses in supply chain and analyzes clinical notes plus estimated patient acuity to improve surgical scheduling and operating-room allocation. For clinical discovery, the hospital's "co-pilot geneticist" combines genetic data, phenotypic information, and global medical literature so physicians can reason through rare-disease cases that had previously gone unresolved. The platform lets use cases that once required long development cycles reach production in days while supporting care delivery, research, and administration at enterprise scale.
Data flow
Operational data such as invoices, clinical notes, and scheduling context flows through the enterprise AI layer into the secure internal ChatGPT environment, where automations route supply-chain work and improve operating-room planning. For rare-disease cases, genetic data, phenotypic information, and global medical literature are brought together in the co-pilot geneticist so clinicians can synthesize evidence and produce diagnosis support more quickly.
Solution architecture
3 components · 3 layers - Compute
- co-pilot geneticist Combines genetic data, phenotypic information, and medical literature to support rare-disease diagnosis and discovery.
- Serving
- secure internal ChatGPT environment Provides a shared, governed workspace where clinical, research, and administrative teams use AI with internal data and workflow context.
- Orchestration
- enterprise AI layer Acts as the hospital's common foundation for deploying new AI workflows quickly with safety, monitoring, and evaluation built in.
- Governance structures were built alongside the technology
- Invoice intake, routing, and responses automated in supply chain operations
- Shared enterprise AI layer instead of one-off tools
- Surgical scheduling uses clinical notes and estimated patient acuity
- Tools that once required extended development cycles can now be deployed in days
"Across more than 50 automations, Boston Children's has captured about 60,000 hours in time savings, which is equivalent to more than $7 million in redeployed labor."
"As a result of this work, more than 40 diagnoses have been made to date that were previously thought impossible."
"The hospital developed what it describes as a 'co-pilot geneticist,' designed to integrate genetic data, phenotypic information and global medical literature."
"The hospital shifted to building what Brownstein calls an enterprise AI layer: a secure internal ChatGPT environment used across research, clinical, and administrative teams."
"Today, more than one-third of employees use AI as part of their daily work."
