R1 RCM AI review (Oct 2026): enterprise revenue-cycle automation, intelligent automation, and human-in-the-loop billing ops

What this R1 page covers

This page is for a revenue-cycle director, patient financial services lead, coding manager, utilization management lead, or CFO at a health system using or considering R1 — as an end-to-end partner or for individual modules. It is not medical, billing, coding, or legal advice.

R1 says it partners with 95 of the top 100 U.S. health systems and has more than $77 billion in net patient revenue under management (home page). Its Phare OS page describes a shared foundation — Phare Intelligence (AI agents and models that read medical records for medical necessity, codes, and appeals), Payer Atlas (1,500+ payer connections), and a Data Platform — under the Access, Claim, and Flow modules. Its applied AI lab, R37, works on payer policy monitoring, autonomous coding, contract analysis, denial prevention, and an AI contact center. These are R1’s descriptions and figures.

Ground rules: the revenue-cycle director owns every automation that can submit, correct, or appeal a claim; the billing supervisor samples what automation touched; the coding manager owns coding changes; the utilization management lead and physician advisors decide patient status; and R1’s figures are its claims, not our measurement.

Major features (with how someone uses them)

Examples at a glance: each feature has a concrete example from a real job plus a short Try this you can run this week.

1. Phare Claim fixes and Phare Flow follow-up on claims (primary)

What it is: On the mid-cycle side, Phare Claim says it will “address claim errors automatically where possible” before submission. On the back end, Phare Flow uses agents that triage, orchestrate, and resolve accounts — “check status, appeal denial, reconcile payments, call payer” — and R1 reports about an 80% cut in manual processing time and more than 40% of AR defects resolved without human involvement. R1’s figures.

“Without human involvement” is the number to sample. A claim fixed or resubmitted by automation is still your claim.

Concrete example (billing supervisor at a three-hospital system) The billing supervisor pulls fifty claims R1’s automation touched last month and sorts them: forty-one clean, six needed an edit a person later made, and three were resubmitted with a modifier the payer rejects for that service. She sends the three to R1 and asks the revenue-cycle director to keep that payer’s rule manual until it is fixed.

Try this

Primary workflow:

Sample fifty claims R1's automation touched last month. Have the billing supervisor classify each as clean, needs edit, or should never have auto-submitted before widening rules.

2. Phare Access eligibility and prior authorization

What it is: Phare Access “blends AI and intuitive patient tools to clear patients early on, automating eligibility and insurance verification, securing prior authorizations, and embedding accurate patient data at intake.” R1 reports 96% of authorizations cleared in one day and up to 45% lower patient-access costs — its figures. R1’s newsroom says its acquisition of Humata Health is meant to strengthen prior authorization within its pre-bill architecture.

An authorization marked “cleared” should match what the payer approved — the service, units, and dates.

Concrete example (patient access director) The patient access director checks twenty authorizations Phare Access marked cleared against the payer’s approval. Eighteen match. One approved fewer therapy visits than were scheduled; one approved a different site of care. She adds a units-and-site check before scheduling the full course.

Try this

Match the approval:

Have the patient access director compare twenty authorizations marked cleared with the payer’s approval for service, units, and dates before scheduling the full course of care.

3. Phare Claim pre-bill review for coding and documentation

What it is: Phare Claim connects coding, charge capture, clinical documentation integrity, and utilization management, using AI to “surface risks earlier” before claim submission, with “R1 clinical and revenue cycle experts” adding judgment on complex cases. R1 reports 100% AI review across cases, first-level reviews in under four minutes, and a $250 revenue lift per discharge for Phare Audit — R1’s figures. The R37 lab says it is working toward autonomous coding with a “DRG-V safety net”.

A flagged documentation gap is a question for the physician, not a code change. Track which flags led to queries and which were wrong.

Concrete example (coding manager) The coding manager reviews thirty inpatient cases Phare Claim flagged before billing. Twenty-two flags were useful queries, five were already resolved in a late note, and three suggested a higher-weighted diagnosis the documentation did not support. She sends physician queries for the twenty-two and asks R1 to explain the three.

Try this

Flags become queries:

Have the coding manager review thirty Phare Claim flags before billing and record which led to a physician query, which were already resolved, and which the documentation did not support.

4. Phare UM for patient status and medical necessity (16 Sep 2026)

What it is: R1 announced Phare UM on 16 September 2026. It says the tool “continuously reviews 100% of cases from admission through discharge,” uses AI agents to assess denial likelihood and payer risk through Payer Atlas, and gives utilization management teams prioritized worklists, “keeping human judgment at the center of every decision.” The Guthrie Clinic is named among the first adopters. In one customer sample assessment R1 says about 65% of observation encounters should have been inpatient — R1’s figure.

Status is a clinical decision made by the physician and the utilization review process, not by a payer-risk model. Use the worklist to prioritize, not to decide.

Concrete example (utilization management lead) The utilization management lead reviews twenty cases Phare UM ranked highest-risk with a physician advisor. Fourteen warranted a status review; four were already documented well; two were ranked high only because of the payer. They keep the worklist for ordering and record every status decision with the physician advisor’s name.

Try this

Rank, don’t decide:

Have the utilization management lead and a physician advisor review the twenty highest-risk Phare UM cases each week and record who made every status decision.

5. Command Center and what may run without a person

What it is: Phare OS includes a Command Center with “unified analytics, role-based controls, and real-time visibility into human and AI operations,” which R1 says lets revenue-cycle managers step in and fine-tune the operating model. R1 says Phare is “built on R1’s AI governance framework” with “expert human oversight and supervision” (r1rcm.com).

Visibility is not the same as a rule. Write down which actions may run without a person and who can change that list.

Concrete example (revenue-cycle director at a regional health system) The revenue-cycle director asks R1 for the list of actions agents take on the system’s accounts without a person — status checks, payment reconciliation, appeals, payer calls — and marks each as allowed, allowed under a dollar limit, or needs approval. Appeals over $5,000 move to the denials manager’s queue.

Try this

Write the list:

Have the revenue-cycle director write down every action R1’s agents may take without a person, with a dollar limit and a named approver for the rest, and review it each quarter.

6. AI-prepared denial appeals

What it is: R1 says Phare Intelligence agents interpret medical records to produce “airtight appeals” (Phare OS page), Phare Flow can “appeal denial” as one of its actions, and the R37 lab is working to automate the workflow “from denial receipt to appeal submission.” R37 also describes R1-super, a reasoning model built with NVIDIA for claims adjudication. These are R1’s descriptions.

An appeal is a statement to the payer on the hospital’s behalf. Someone at the hospital should be able to defend every sentence.

Concrete example (denials manager) The denials manager reads ten AI-prepared clinical appeals before they go out. Eight are accurate and cite the right records. One quotes a lab value from the wrong day; one cites an outdated payer policy. He fixes both and asks R1 to hold clinical appeals over $5,000 for his review.

Try this

Read before submitting:

Have the denials manager read ten AI-prepared clinical appeals each month before submission and check every cited record and payer policy.

Pricing context (October 2026)

R1 publishes no price list for Phare OS, its modules, or its end-to-end services. When we checked on 19 Oct 2026, r1rcm.com/pricing returned a not-found page; R1 offers enterprise partnerships and individual solutions through its sales team. Treat it as quote-only.

Ask in writing whether each module is priced per transaction, per claim, as a percentage of collections, or as part of a broader operating partnership, and how fees change when automation replaces work R1 staff did before.

Uncertainty: We did not test R1. Authorization, processing-time, AR-defect, revenue-lift, and status figures are R1’s claims. Phare UM was announced on 16 September 2026; ask which modules are generally available for your setting. Confirm BAA terms, which actions agents take without a person by default, and how you can export the audit trail.

When it’s useful (honest take)

Useful when you are a hospital or health system with large claim volumes, considering an end-to-end revenue-cycle partner or module-level automation, and the revenue-cycle director will own every rule that submits or appeals.

Less useful when you are an independent practice that needs PM/EHR software with RCM services (AdvancedMD AI or athenahealth AI), a clearinghouse-led platform (Waystar AI), or autonomous coding alone (Akasa AI).

Related options (educational)

FAQ

What is R1 Phare OS?

Phare OS is R1’s name for its revenue-cycle platform: AI agents and people working across Phare Access (eligibility, authorizations, intake), Phare Claim (coding, documentation, pre-bill review), and Phare Flow (claim follow-up, denials, payments), on a shared foundation of Phare Intelligence, Payer Atlas, and a Data Platform.

What is Phare UM?

A utilization management module R1 announced on 16 September 2026. R1 says it reviews every case from admission through discharge, assesses denial likelihood and payer risk, and gives teams prioritized worklists, with human judgment at the center of each decision.

How much does R1 cost in October 2026?

R1 publishes no prices; its pricing URL returned a not-found page when we checked on 19 Oct 2026. Ask for a written quote and how fees are calculated.

Can R1’s agents resubmit or appeal claims without a person?

R1 describes Phare Flow agents that check status, appeal denials, reconcile payments, and call payers, and reports more than 40% of AR defects resolved without human involvement. Decide which actions may run without a person and sample the results monthly.

Who should check R1 AI output?

The billing supervisor for claims automation touched; the patient access director for authorizations; the coding manager for pre-bill flags; the utilization management lead and physician advisors for status; the denials manager for appeals; and the revenue-cycle director for the list of automatic actions.

Optional next step (if seats overlap)

If you pay for R1 services, a separate clearinghouse, a coding AI, and a patient billing vendor that all touch the same claims, that overlap is worth mapping. Map the overlap with the $19 AI toolkit audit or the $9 AI toolkit spreadsheet. Email Audit $19 or Toolkit spreadsheet $9.

Checkout is live via Stripe. After you pay, we email delivery (spreadsheet) or intake next steps (audit).

Related: athenahealth AI · Epic Systems EHR AI · Waystar AI · Oracle Health AI · Abridge AI · Microsoft Dragon Copilot AI · Suki AI · athenahealth vs Epic Systems EHR · eClinicalWorks AI · Akasa AI · Phreesia AI · Oracle Health vs athenahealth · NextGen Healthcare AI · Elation Health AI · Innovaccer AI · Notable AI · Cedar AI · Luma Health AI · Abridge vs Microsoft Dragon Copilot · Waystar vs Akasa · ChatGPT for healthcare ops · Gemini for healthcare ops · What’s new in athenahealth (Sep 2026) · What’s new in Epic Systems EHR (Sep 2026) · AdvancedMD AI · Stack map for healthcare ops · Applied Epic AI (insurance AMS — different product) · Microsoft 365 Copilot · ChatGPT for finance · Claude · Three checks before another AI seat · How-to: Turn on R1 Phare Flow for claim follow-up and denials, and have a revenue-cycle supervisor approve every appeal · ops hub