Read the Report — State of Applied AI →
HealthcareProduct Development

How CirrusMD Uses Writer to Achieve 234% More Benefits Referrals

CirrusMD, a physician-first virtual care company connecting over 13 million members with physicians in under 60 seconds, deployed Writer’s Palmyra Med model to solve two critical challenges: helping physicians recommend the right patient benefits at the point of care, and automating clinical documentation. After a year of failed in-house development with OpenAI models, CirrusMD launched its first AI-powered workflow in under six months with Writer. The result: a 234% increase in physician direction to benefits and 15x higher benefits engagement by patients.

Impact

234%

Increase in physician direction to benefits

15x

Increase in patient benefits engagement

Hundreds of thousands

Patient interactions per year through AI workflows

<6 months

Time to launch first AI use case with Writer

Challenge

CirrusMD needed AI that could surface patient-specific benefits recommendations to physicians in real time and automate clinical documentation, but a year of in-house development with OpenAI models produced inconsistent results and delayed its first launch by months due to model instability and limited engineering resources.

Solution

CirrusMD deployed Writer’s Palmyra Med model and full-stack platform to build benefits navigation and clinical documentation workflows, launching the first use case in under six months and completing a second in parallel—without expanding its AI engineering team.

Tools & Technologies

What Leaders Say

Within just six months of working with WRITER, we launched our first use case and completed the technical build for a second. This is a huge leap forward compared to the year we spent working toward an in-house solution.

Tanya Dillard, SVP of Product Management and Innovation, CirrusMD

WRITER’s Palmyra Med model is trained specifically for medical use cases, which means it delivers results we can rely on.

Tanya Dillard, SVP of Product Management and Innovation, CirrusMD
Get the full context.

Sign up to read complete case studies, access detailed metrics, and unlock all use cases.

Full Story

CirrusMD sits at the intersection of patients, physicians, and plan sponsors across commercial and government health plans. Its chat-based care model connects over 13 million members with a physician in under 60 seconds—handling not just acute care but also behavioral health, chronic condition management, and primary care. Each interaction touches three parties with competing pressures: plan sponsors want members maximizing their coverage, physicians are time-constrained and information-limited, and patients face access barriers driven by cost, distance, and complexity.

The opportunity CirrusMD identified was precise: physicians who personally recommend benefits to patients drive far higher engagement than any other channel. But identifying the right benefits for a given patient—tailored to their health situation, coverage plan, age, and location—during a live care interaction was impractical without AI. At the same time, clinical documentation consumed physician time that should go to patients. CirrusMD spent a year attempting to build these capabilities in-house, cycling through GPT-3, GPT-3.5 Turbo, and GPT-4. Each model update required rebuilding pipelines from scratch. Output quality was inconsistent, and the models lacked the medical specificity required for clinical workflows. With just one ML engineer, the team was stretched thin and months behind on its first launch.

CirrusMD switched to Writer and its Palmyra Med model, a large language model trained specifically for medical language and clinical context. The Writer full-stack platform provided all the components needed to build and deploy AI-powered workflows without stitching together disparate tools or scaling an internal AI team. Writer’s services team met CirrusMD’s tight development deadlines. The first use case—real-time, patient-specific benefits recommendations surfaced to physicians at the point of care—launched in under six months. A second use case, clinical documentation automation, was completed in parallel.

The impact on benefits engagement was dramatic. Physician-directed benefits referrals increased by 234%, and patient benefits engagement rose 15x. Hundreds of thousands of patient interactions per year now flow through Writer-powered AI workflows. Physicians report that the benefits navigation feature delivers real value to patients, and plan sponsors see measurable improvement in benefits utilization—the metric they care most about.

CirrusMD’s roadmap extends the AI layer across more care categories and use cases. The shift from in-house development to Writer as a platform partner gives the company the speed and clinical reliability to expand without scaling a dedicated AI engineering function. The lesson from the transition was direct: building AI infrastructure is not CirrusMD’s core business—delivering quality care is, and Palmyra Med makes that possible at scale.

Similar Cases

V
Vizient
4x
roi vs. initial estimate

Vizient, the largest member-driven healthcare performance improvement company in the United States serving over half of all US healthcare organizations, deployed Writer to help more than 100 collaborators produce persona-specific, compliance-ready marketing content at scale. With automated brand and regulatory guardrails built into every workspace, the team created a 60-page flagship industry report with fewer editing cycles and launched persona-based content personalization that the team previously lacked the staff to attempt. In its first year, Writer delivered 4x the expected ROI and $700,000 in savings.

HealthcareWWriter
EH
Elation Health
61%
reduction in time to first insight

Elation Health migrated its Clinical Insights feature to Claude, achieving a 61% reduction in time-to-first-insight for chart review and doubling adoption among clinicians. The platform serves 46,000+ clinical users across 50 states, helping primary care physicians synthesize dense patient histories before appointments.

HealthcareCClaude
H
Humana
~66% (1/3 cost)
cost reduction

Humana replaced its legacy IVR system with an IBM Watson-based conversational voice agent that handles 7,000+ provider calls daily, completing inquiries in 2 minutes at one-third the previous cost.

HealthcareICIBM CloudIWIBM watsonx Assistant
M
Mediq
55
active automations in production

Mediq, an international healthcare company operating across 14 European countries, formalized a group-wide Center of Excellence for automation in 2024 built on UiPath. By year-end the CoE ran 55 automations saving 55,000 hours annually, with UiPath Document Understanding processing sales orders at 98% accuracy across regulated healthcare supply chains.

HealthcareUMUiPath MaestroUPUiPath Platform
A
ArisGlobal
diagnostic file collection time reduced from 45–60 minutes to near-instant

ArisGlobal, an AI-first life sciences software company serving global pharmaceutical organizations, deployed Datadog APM, App Builder, Workflow Automation, and On-Call to enhance observability and automate operations for its LifeSphere platform. Automated incident remediation cut diagnostic file collection from 45–60 minutes to near-instant, and the team achieved 100% automation of previously manual operational tasks.

HealthcareDDatadog
M
Medlitix
90% (70min to 6min)
review time reduction

Medlitix implemented UiPath medical record summarization with DeepRAG, cutting clinical review from 70 minutes to 6 minutes per case (90% faster) with 95% accuracy and $1.2M savings.

HealthcareUPUiPath Platform
D
Doctolib
Weeks to hours
technical debt migration time

Doctolib deployed Claude Code across its entire engineering team after a 30-engineer pilot, cutting codebase onboarding from weeks to days and compressing a full technical debt migration from weeks to hours. Engineers now get instant pull request reviews without waiting for team availability, freeing them to focus on healthcare feature development.

HealthcareCCClaude Code
UM
UChicago Medicine
100
patient marketing segments built

UChicago Medicine deployed Salesforce Agentforce for Health alongside Health Cloud and Marketing Cloud to provide 24/7 AI-driven patient support. Autonomous agents handle prescription refills, appointment scheduling, insurance verification, and parking logistics — tasks that previously required live staff and averaged 4 minutes each. The team built 100 targeted patient marketing segments in just 4 months.

HealthcareSASalesforce AgentforceSHSalesforce Health Cloud