See how physician time is actually spent - and defend it in an audit.
RealActivity replaces manual time studies with a continuous, CMS‑accepted methodology that makes physician activity visible, auditable, and actionable - by the 15th of every month. Governed AI, on the provider side and the patient side.
Memory is not data.
For decades, physician effort has run on the least trustworthy data there is - human memory, reconstructed weeks after the fact and stretched across a year.
RealActivity doesn't ask anyone to remember. It captures activity as it happens and has a clinician confirm it in under thirty seconds. The result is a record you can do two things with that no one else can: defend it in a CMS audit, and safely govern an AI on it.
Capture at the source
Activity recorded as it happens, across the full period - no recall, no sampling.
Provider attests - under 30 seconds
The clinician reviews their own AI-curated record and signs off. A named human verifier on every entry.
Governed record
An attested, auditable record with provenance - ready to defend, and safe to feed an AI system.
Governed agents on both sides of care.
Governed data and a governed provider assistant on one side. Governed patient agents on the other. The human-verified, auditable record is the bridge between them. The RealActivity team is partnering with Anthropic (in progress) to extend governed Claude Mythos Preview access for defensive review of hospital code and systems - on the same OpenClaw harness as Tula and Aria.
Provider Activity & Ava
The system of record for how physician time is allocated - and an assistant that helps providers verify it.
- Provider Activity Platform - CMS-accepted continuous attestation, by the 15th.
- Provider Activity Studio - the workspace where providers review and attest.
- Ava - the Activity Verification Assistant, governed by design.
Tula & Aria
Open-source skills patients use today - records, visit prep, and Lookout for me - plus the governed platform hospitals run at scale.
- health-records - pull medical records via SMART on FHIR
- prep-my-visit and Lookout for me (myhealth-pulse)
- Aria - governed patient agents at hospital scale
Microsoft made work agents board-ready. The patient chart still needs a governance plan.
Aria is RealActivity's governed patient-agent layer for health systems, built on open-source Tula and the OpenClaw runtime Microsoft spotlighted at Build. In a 15-minute briefing, the team shows a live MyChart pull, visit prep, Lookout for me, and the audit trail around every model call.
See the Build26 briefing page Join My Aria waitlist Watch the demo
One platform. Three steps. Data you can defend.
AI curates activity data
RealActivity integrates with your ERP and EHR to compile all physician activity - clinical, research, teaching, and administrative - into a single, system-wide view. No manual collection. No surveys.
Providers attest in 30 seconds
Each month, providers review their AI-curated data and attest to their own percentage allocation. The form takes under 30 seconds - 60% are completed within 4 hours of notification.
Leadership gets defensible data
By the 15th, you have 100% accurate, provider-verified activity data. One dataset powers compliance, compensation, workforce planning, benchmarking, and audit preparation.
Meet Ava. Governed by design.
Ava is the agentic assistant inside Provider Activity Studio, built on the same OpenClaw runtime that powers Tula. Providers ask questions about their activity data and complete the monthly attestation conversationally - in under thirty seconds.
Grounded in the attested record
Ava answers from the provenance-rich activity record - not from guesswork.
Human in the loop
The provider attests. Ava assists - it never decides the allocation.
Aligned with the NIST AI RMF
Designed to align with the NIST AI Risk Management Framework and its Generative AI Profile.
HIPAA, with a BAA
Operates under HIPAA with a Business Associate Agreement for deployments that touch PHI.
15+ years in production at a leading academic medical center.
The methodology was developed in direct discussions with CMS as an alternative to traditional time studies, and has supported successful third-party audits.
Defensible where it counts
- Medicare Cost Report (Schedule J) built on defensible physician time data
- 340B drug program eligibility documentation maintained automatically
- Successful third-party audit preparation with HHS and CMS
- Compensation models defensible with RVU-matched benchmarks
Capacity you can see
- Clinical Workforce Committee makes hiring decisions backed by FTE data
- Surplus capacity identified and redeployed across affiliate locations
- Grant expiration planning gives months of advance notice
- New affiliates onboarded within 2 months
// Phrasing note: "accepted by CMS as an alternative methodology." Case-study quotes pending Dartmouth sign-off before publication.
The pressure on physician effort data is rising fast.
Unprecedented CMS audit expansion
CMS has expanded audit activity from roughly 60 to 550 plans annually, with medical coding staff scaling from 40 to 2,000.
Physician workforce crisis
The U.S. faces a projected shortage of 141,000+ physicians by 2038. AMCs that can't optimize capacity face compounding shortfalls.
340B under scrutiny
The 340B program faces intensified HRSA audits and manufacturer restrictions. Accurate effort documentation is foundational to compliance.
Rising physician labor costs
Physician labor costs rose more than $42 billion between 2021 and 2023. Leaders need precise data on how that investment translates to output.
Ready to make physician activity visible?
We offer a structured 90-day pilot on your data, with your providers, in your environment. The best first conversation connects your CMO and Compliance leadership.
hi@realactivity.com · +1 857-206-1190 · Gloucester, MA