One of the largest outpatient diagnostic imaging providers in the U.S. set out to modernize a critical part of its operation: how diagnostic test appointments are scheduled. Originally, the organization planned a traditional UI modernization. But faced with the complexity of scheduling, they realized a simple redesign wasn’t enough – and accelerated straight to an AI-driven solution.
They chose to start with mammograms, where the existing scheduling workflow was very complex, slow, and prone to error. The scheduling technician needed to select the right procedure code from over 300+ options based on the patient history and physician notes, using a complex set of rules.
A mistake could mean restarting the process – or worse, booking the wrong appointment, leading to delayed care, wasted clinical capacity, and poorer patient outcomes. It can also hurt the patient experience and reduce retention, as dissatisfied patients may not return.
With PromptQL, the customer is building an AI-powered, scheduling co-pilot that improves operational efficiency, enables faster and more accurate scheduling, and supports better patient outcomes. Once fully rolled out across the company’s hundreds of locations in North America, the customer expects productivity gains of up to $50 million.
"Up to $50m in savings - just by reducing scheduling time and improving resource utilization."
The Challenge
When patients called in to schedule critical diagnostics tests such as mammograms, the scheduler had to work with a legacy, 20-year old RIS (Radiology Information System) that forced them to manually navigate across dozens of tabs, memorize business rules, and select the appropriate procedure codes from a list of more than 300 options.
This task alone – selecting the right procedure code based on a patient's age, history, symptoms, and logistical needs – could take up to 12 minutes per appointment. The cost of mistakes was high: incorrect codes meant canceled appointments, rework, lost revenue because referring physicians are unhappy, and most importantly, patient dissatisfaction and delayed diagnostic care.
In the customer’s own words: “We just lost valuable time because it was scheduled incorrectly. And now we have an angry patient who's complaining to their referring physician, who in turn might decide to not send their patients to us anymore.”
Training new schedulers also posed a significant challenge. It required weeks of onboarding and made scaling the contact center difficult and expensive – especially in high-cost labor markets.
“The UI is very complex, there are tons of business rules, and hundreds of procedure codes. It requires months of training before a technician becomes proficient, fast, and accurate."
How PromptQL Solved It
With PromptQL, the customer was able to convert this complex, manual process into a highly reliable scheduling copilot. Specifically, PromptQL creates value for the diagnostic company and patients in three different areas – and unlocks a fourth one for the future:
- Automatic procedure code selection: Given inputs like patient age, symptoms, or mobility needs, PromptQL recommends the correct procedure codes in real time with high accuracy.
- Smart appointment slot suggestions: Based on availability, interpreter requirements, where prior records are located, or other special accommodations such as a wheelchair.
- Smart form rendering: PromptQL's AI dynamically generates follow-up questions based on initial patient inputs.
- (Potential future enhancement) Voice-to-form automation: As an optional future enhancement, PromptQL could also transcribe live scheduler-patient phone calls and autofill form fields - which reduces both effort, and errors.
“Instead of the scheduler selection from 300 options, PromptQL can help us narrow down the procedure code - quickly and accurately!"
Business Outcomes
Reducing the time it takes to schedule one appointment – just the procedure selection step - by half unlocks significant efficiencies. Each scheduler can handle more appointments per shift – and the customer can mitigate the impact of costly training, staffing, and labor overhead.
But a much bigger opportunity lies in scale.
“Every extra minute spent by a scheduler has a real cost. Reducing just that one 12-minute step unlocks more volume – and makes every call center more efficient."
According to the customer’s internal estimates, once this approach is rolled out across the customer and its broader customer base, it could represent a $50M+ business impact through a combination of:
- Increased throughput (more patients scheduled every day)
- Lower training and staffing costs (faster ramp-up for new staff)
- Fewer appointment errors and reschedules (fewer cancellations and/or rework)
- Higher patient satisfaction and referral retention (better retention and referrals)
Going forward, PromptQL also streamlines ongoing maintenance – generating rule-based workflows dynamically instead of requiring weeks of developer time to hardcode new procedures, codes, or scenarios.
What’s Next
With PromptQL, the customer is building scalable, explainable, and efficient workflows that reduce friction for patients, schedulers, and developers alike.
Following the success of the mammogram use case, the customer is now preparing to expand the model to other diagnostic imaging workflows such as CT scans, X-rays, and PET exams. The long-term goal is to build a generalizable, AI-assisted scheduling platform that the customer can offer commercially to other diagnostic lab operators.
With PromptQL, the future of intelligent healthcare operations is already here.
