Luma Health Accelerates the Shift to Operational AI Across Health System Workflows
Health systems are stretched thin-labor shortages, fragmented processes, and rising demand keep piling on. Luma Health is pushing past one-off automation with Operational AI that runs workflows end to end across the EHR and downstream systems. In 2025 alone, customers saved 2MM+ staff hours; to date, that total exceeds 2.5MM hours across 50+ health systems.
The platform is powered by Spark, Luma's AI core, which coordinates agents and workflows across access, engagement, intake, and payment capture. The goal is straightforward: reduce manual work, recover capacity, and create predictability in day-to-day operations.
From Point Tools to System-Level Execution
Luma's Operational AI spans four interconnected categories of work. Instead of optimizing a single task, it carries work from start to finish with context shared across systems.
- Access: New patient acquisition, referral and fax workflows, order orchestration, scheduling and rescheduling, and waitlist management.
- Engagement: Conversational patient access, intelligent messaging, and patient-initiated interactions across channels.
- Intake: Responsive intake and registration, urgent care wait time visibility, clinic flow optimization, and readiness checklists for patients and staff.
- Payment Capture: Eligibility verification, prior authorization support, and copay and balance collection.
What Operations Leaders Are Seeing
- 2.5MM+ staff hours saved to date (with 2MM+ in 2025 alone).
- 350,000+ care-related next steps handled.
- Hundreds of agentic workflows live, including intelligent no-show recapture, post-discharge follow-up, and closed-loop referral management.
- More revenue realized by bringing in more demand, filling more slots, collecting more payments, and converting more referrals.
These outcomes span leading EHR environments-Epic, MEDITECH, eClinicalWorks, Oracle Health, and athenahealth-where predictable throughput matters.
Document Flow Agent and Order Management Agent: Intelligent Fax, Order, and Referral Orchestration
Fax-based workflows slow everything down. Luma's Document Flow Agent automates referral intake, prescription refill requests, and ROI processing at organizations such as the University of Arkansas for Medical Sciences, Northfield Hospital + Clinics, and DENT Neurologic Institute-no manual sorting, tracking, or follow-up required. Once a referral is processed, Spark triggers downstream tasks like patient outreach and self-service scheduling, then closes the loop with referring providers.
"Faxes often create bottlenecks that require staff working manually to avoid delays in care, and at a large specialty organization like DENT the volume is unsustainable," said Emily Smythe, EMR & Technology Systems Manager. "Luma's AI is fast, accurate, and gets more patients in the door."
Conversational Agent: Omnichannel Patient Communication
High call volumes and long waits burn staff time and delay access. Luma's Conversational Agent works as an omnichannel concierge and hands off to the Scheduling Agent to book directly in the EHR-no manual callbacks.
At Sun River Health, the agent now manages a large share of inbound access interactions and only routes to staff when needed. Call abandonment fell by more than one-third, and average answer times dropped 75 percent. "We've partnered with Luma for years to help patients connect with us more easily, no matter their needs," said Carlos Ortiz, COO at Sun River Health. "Luma's AI agent now handles a large volume of access work autonomously, helping our call center serve patients faster without adding staff."
Agentic Innovation Across the Platform
As systems consolidate vendors and push for measurable outcomes, Luma continues to expand Spark-orchestrated workflows across outbound engagement, financial clearance, prior authorization, and revenue recovery. Many customers now design and govern their own multi-agent workflows in Luma, extending Operational AI across teams while maintaining controls.
"For a decade, we've used the best available technology to prevent digital dead ends and move patients to the next step," said Marcelo Oliveira, Luma's Chief Product and Technology Officer. "AI now enables the orchestration that healthcare access needs-proactive, connected, and informed by EHR data."
Near-Term Roadmap
- Outbound conversational coordination: The Conversational Agent will proactively follow up by reviewing EHR data, visit notes, and prescriptions to identify incomplete next steps and reduce missed follow-ups.
- Prior authorization coordination: The Order Management Agent will manage end to end-from eligibility checks and payer outreach to scheduling after approval-to shorten clearance timelines.
- AI workflow builder: Operations teams will create and govern multi-step AI workflows directly in Luma without custom development or third-party tools.
Why This Matters for Operations
End-to-end orchestration changes how work gets done: fewer touchpoints, fewer handoffs, and fewer stalls between systems. The impact shows up as recovered staff time, faster throughput, and cleaner revenue.
- Start with two or three high-friction workflows (faxed referrals, prior auth, no-show recapture) and define a clear "start-to-done" path.
- Instrument baseline metrics-hours spent, turnaround time, abandonment, and slot utilization-then track weekly deltas.
- Prioritize EHR-integrated execution so tasks complete without manual callbacks or swivel-chair steps.
- Set guardrails: clear KPIs, exception queues, and human-in-the-loop thresholds for edge cases.
- Tie outcomes to capacity and revenue: appointments filled, days-to-schedule, collections, and referral conversion.
Context: Prior Authorization Pressure
Prior authorization remains a major operational drag. For policy context and timelines that affect your clearance workflows, see CMS's Interoperability and Prior Authorization Final Rule fact sheet.
Upskill Your Team
If you're building internal capabilities around AI-driven workflow design and governance, explore AI for Operations for practical training paths.
Learn More About Luma's Operational AI
- Visit Luma's website.
- Visit Booth #832 at ViVE.
- Attend ViVE session "How AI Fixed a Hidden Fax Drain on Staff Time," with Luma co-founder and president Aditya Bansod and Northfield's Debbie Oathoudt.
About Luma Health
Luma Health's Operational AI platform removes bottlenecks to make health systems more efficient-from patient-facing touchpoints to the back office. Headquartered in San Francisco with R&D in the United States, Brazil, and Europe, Luma serves more than 1,000 healthcare organizations across the United States, Canada, the United Kingdom, and the Caribbean. Luma supports the care experiences of more than 100 million patients.
Media Contact: lumahealth@nextpr.com
Your membership also unlocks:
