HCA's Palantir-backed Timpani software is disrupting nurses' shifts

Summary

Timpani, a shift scheduling tool developed by Palantir and deployed at approximately 130 of HCA Healthcare's 190 US locations since 2023, is facing continuous complaints from nurses. Nurses report that the tool ignores shift preferences, understaffs teams, and assigns inexperienced personnel. Meanwhile, a separate Palantir Foundry tool used at Rayus Radiology is alleged to have linked orders to incorrect patient profiles.

Why it matters

This represents one of the clearest real-world cases of an AI-based shift scheduling system degrading frontline healthcare working conditions at scale. Nurses characterize the system's output not merely as a nuisance, but as a patient safety risk. The parallel lawsuit, which alleges data deletion to hinder independent audits, alongside Palantir's claim that they 'only provide information,' highlights a gap in oversight mechanisms. This situation could set a precedent for integrating AI into other regulated sectors with minimal external oversight.

Highlights

• Nurses report that Timpani routinely assigns consecutive 12-hour shifts, keeps the number of senior nurses insufficient, particularly on Sundays, and invalidates stated leave requests far more frequently than the previous manual process.
• A former HCA data science manager claims they were fired for raising concerns about Timpani and that the company deleted shift data to prevent independent audits of the algorithm's performance.
• The Palantir Foundry tool used by Rayus Radiology produced scheduling errors involving incorrect patient links and fabricated personal information, indicating that the platform's reliability issues extend from nursing to clinical imaging workflows.

Critical take

The article relies heavily on anonymous nurse statements and a single, unresolved lawsuit. HCA's own metrics (a 1% red day rate and a 98% skill mix claim) are presented without independent verification. The lack of third-party audits of Timpani's actual shift outputs leaves both the nurses' complaints and HCA's efficiency claims as untestable assumptions.