The New Front Door: How Agentic AI Powers Patient Access at Clear Medical Imaging

At any given moment, hundreds of referrals are in motion across Clear Medical Imaging (CMI). They arrive by fax, by email, in-person and through the Ocean e-referral system — typed and handwritten, in different templates and formats. Every one needs to be read, validated, matched to a patient and turned into a booked appointment.

For Michael Reinkober, CEO of CMI, the math behind that operation was straightforward: every increase in referral volume meant adding more staff to do more of the same manual, repetitive tasks: file renaming, requisition parsing, slot-checking, outbound dialing. Timely, high-quality patient care is at the heart of CMI’s mission, and meeting that bar at scale meant finding a system that could improve patient access while simplifying operations. PocketHealth Conductor delivered both.

The Challenge

Every step of CMI’s intake depended on people — and growth meant adding more of them.

CMI’s 20-person intake operation was sized to match referral volume. Referrals arrived by fax, email, walk-in and the Ocean e-referral system, in formats ranging from typed PDFs to handwritten orders. Staff members renamed the files, validated the requisition, looked up available slots and called the patient. Every handoff between tasks depended on someone being awake, at a desk, with capacity to take the next step.

For Michael, the trajectory was clear: scale the team linearly with referral volume, or move the repetitive work somewhere else and free up capacity for the patient care that actually needs a human. What CMI needed was a system that could promise speed of care and improve patient access by:

• Reading every incoming referral the moment it arrives, regardless of channel or format
• Reaching the patient immediately, while the conversation with their referring physician is still fresh
• Booking the appointment end-to-end without staff making outbound calls
• Surfacing every exception clearly, with full context, so staff can resolve cases that genuinely need a human
• Working across every modality and every CMI location, from day one

That’s what pushed CMI to deploy Conductor, PocketHealth’s agentic AI platform for radiology’s back-office.

The Solution

Conductor automates CMI’s full referral-to-booking workflow — improving patient access and simplifying operations.

Conductor didn’t ask CMI to rip out any of its current systems. It slotted into the existing workflow, starting at the network folder where incoming referrals were already being collected. Conductor monitors that folder, picking up every new file, parsing it and kicking off patient outreach.

The parsing pipeline is powered by an LLM with OCR grounding. Referrals arrive in many formats; each has to be recognized, cropped, cleaned and read accurately enough to book a patient without introducing errors. Each field — patient demographics, referring physician, procedures ordered — gets a confidence score. Anything below a pre-set threshold flags for human review rather than passing through. That’s the primary guardrail against automation errors: if Conductor isn’t sure, a human sees it before anything moves.

The next step is patient outreach, which runs as a coordinated voice and SMS flow. Conductor calls the patient first with a short voice message letting them know CMI is ready to book their appointment, then follows with an SMS to handle the scheduling itself. From there, it’s a live two-way conversation. Patient responses are captured, structured and fed back into the automation in real time, driving slot selection, booking the appointment directly in CMI’s RIS, and delivering modality-specific prep instructions to the patient. The voice call does the trust-building. The SMS makes booking effortless.

What used to be one of CMI’s most labour-intensive workflows now can run at a fraction of the cost. The volume that once depended on a 20-person intake operation now moves through Conductor. And CMI’s operational investment is reinvested where it matters most: in direct patient care.

The Results

Patients are rating the booking experience 4.5 out of 5, and CMI’s team is spending its time where it counts most.

For Michael, the question was never just whether the system ran. It was whether the patient experience was actually better and whether the team was freed up to focus on the work that needed them. On both measures, the numbers moved. Conductor delivers:

• 99%+ parsing accuracy across handwritten and typed incoming referrals. Patient data, referring physician details and ordered procedures are extracted correctly from incoming referrals. No wrong exams booked. No clean data being re-keyed by hand.
• 99%+ order creation accuracy. Orders land in the RIS correctly the first time. No silent data errors downstream.
• 99%+ appointment scheduling accuracy. Bookings arrive at the right location, for the right modality, directly in CMI’s RIS, at a time the patient actually chose.
• 4.5/5 average patient rating. Patients are rating the experience themselves — a direct read on whether CMI’s front door actually feels easier to walk through.

The operational reality behind the numbers: every referral is picked up, parsed and acted on within minutes. Outreach goes out immediately, not the next morning. Patients are getting booked while the conversation with their doctor is still fresh, without anyone on the team needing to dial a single number. And CMI’s team is back where they want to be: with the patients and conversations that actually need them.

About PocketHealth

PocketHealth is the AI-enabled workflow automation platform transforming health system operations. With three interconnected solutions—Conductor, Image Exchange and Patient Connect—PocketHealth automates all manual, non-clinical tasks across the patient journey, from requisition processing to inbound call management to seamless image sharing and patient engagement. As a result, providers can redeploy staff, expand margins and deliver safer, more seamless care, while patients gain the knowledge and guidance they need so follow-ups don’t fall through the cracks. PocketHealth is SOC2- and HIPAA-compliant, and is trusted by more than 900 hospitals and imaging centers across North America. Request a demo today.