A Commercial Tool Left 3,000 Clinicians Without the Data They Needed
NorthStar Anesthesia manages anesthesia staffing across 25+ U.S. states with roughly 3,000 physicians and Certified Registered Nurse Anesthetists. Those clinicians rotate across facilities, overnight shifts, and on-call rotations. Most of them have no steady computer access during a shift. They check their phones between cases.
When NorthStar adopted a new commercial scheduling platform, it handled most functions adequately. But it deliberately hid colleagues' time-off data. For a workforce that trades shifts constantly, that omission made scheduling painful and opaque.
NorthStar and implementation partner Synaptiq had already built a Databricks foundation — scheduling, time tracking, and contract data unified in a medallion architecture. The instinct was to patch the gap with another dashboard. A pilot showed that wouldn't work. The dashboard wasn't mobile-friendly enough. "It just didn't look clean in the way that we wanted it to," said Erin Sarosi Bell, Program Manager at Synaptiq. "We pivoted to the app."
The Existing Stack Made a Fast Build Possible
Synaptiq made the case for building directly on Databricks Apps. The data was already there. Governance was already configured. Microsoft Entra ID SSO could extend to the entire clinical workforce without standing up new infrastructure.
A single software engineer built a React TypeScript app deployed via Databricks Apps in only a few weeks. The app includes color-coded shift views by clinician type, facility selectors, day/week/month calendar modes, shift notes, and search by shift type. It refreshes every 30 minutes. Critically, it surfaces the time-off data the commercial platform buried.
"Within a matter of weeks, we were able to build something that met the need and iterate on it rapidly, pushing out multiple releases," said Dan Levine, CTO at NorthStar. "It solved a huge pain point for our users."
Daily unique users climbed from 75-80 at launch to 110+ as the first wave of clinicians migrated. The source reports "dozens and dozens" of users writing in to say the app changed how they work and reduced scheduling stress.
What This Means for Teams Evaluating Workforce App Development
The NorthStar story, as reported by Databricks, is a vendor case study. The source does not provide independent cost figures, development timelines from other vendors, or comparable benchmarks. Readers should weigh it accordingly.
That said, the operational logic is worth examining on its own terms.
The speed came from one specific condition. The data platform already held unified, governed data. Identity infrastructure was already in place. Without those prerequisites, the same build would have required integration work, new credential management, and governance decisions that could easily extend the timeline.
Metrotechs analysis: the decision facing CIOs and data leaders is not whether to build custom workforce apps. It is whether their current data platform eliminates or creates that prerequisite work. If your scheduling, time tracking, and workforce contract data live in separate systems with inconsistent governance, a fast build on any platform becomes a slow integration project first.
NorthStar's next planned additions include AI/BI Genie for natural language shift queries, push notifications, and automating the morning staffing report with Databricks. The announcement does not confirm timelines or release status for those features.
The Audit Question for Your Team
Before starting a similar project, ask: does your data platform already hold the relevant workforce data in a governed, unified layer? Is enterprise SSO already extended to your end-user population? Can a single engineer deploy a front-end app without opening new infrastructure requests?
If any answer is no, factor that work into the actual timeline. The weeks-long build NorthStar achieved was downstream of prior platform investment, not a shortcut around it.