AI booking support across 29 locations. 78% less abandonment.
Deploying AI agents with tool calling across Canada's leading ultrasound brand — from knowledge base to live customer support.
How UC Baby partnered with SundayPyjamas to build an AI support agent that handles package recommendations, location lookups, and gestational timing guidance — serving 85% of customer inquiries without staff intervention.
UC Baby
85%
Inquiries handled by AI
92%
Package recommendation accuracy
78%
Reduction in booking abandonment
29
Locations served nationwide
The problem
UC Baby operates 29 ultrasound clinics across Canada. Expectant families had complex questions about optimal timing for different ultrasound types based on gestational age, package details, location availability, and safety information. Without immediate answers, customers abandoned the booking process, called multiple locations, or booked packages inappropriate for their gestational stage.
Front desk teams across all 29 locations were overwhelmed with repetitive inquiries, reducing their capacity to focus on in-person customer service. The booking abandonment rate was high, and the mismatch between customer questions and available staff time was the core bottleneck.
What we built
We deployed an AI support agent using SundayPyjamas AI Suite with tool calling for package recommendations, location lookups, and contextual booking guidance. The agent pulls from UC Baby's service catalog, pricing data, location information, and gestational timing guidelines to provide accurate, specific responses.
The agent matches ultrasound packages to gestational age, checks location availability by postal code, and guides customers to the right booking action — handling 85% of inquiries without staff intervention. Staff time shifted from answering repetitive questions to delivering in-clinic experiences.
What we built for UC Baby
- AI agent with tool calling for package recommendations, location lookups, and booking guidance
- Knowledge base over service catalog, pricing, location data, and gestational timing guidelines
- Context-aware responses that match ultrasound packages to gestational age
- Multi-location support across 29 clinics with location-specific availability
- Safety boundaries and response guidelines for healthcare consumer interactions
- Analytics dashboard tracking agent performance, conversion, and satisfaction
Technical integration
Tool calling for package matching, location availability, knowledge retrieval over the full service catalog, and contextual booking guidance.
Package recommendation
Tool calling → service catalog lookupgestationalAge + preferences → matched packages
Location lookup
Tool calling → location availabilitypostalCode → nearest locations + availability
Knowledge retrieval
File search over UC Baby knowledge basequery → service catalog + FAQ → grounded response
Booking guidance
Contextual agent responsecustomer intent → package + location + timing → booking CTA
Geographic coverage
The agent serves customers across all Canadian provinces: Ontario (36% of interactions), British Columbia and Saskatchewan (16% each), Maritime provinces (16% combined), and Western and Atlantic regions (12% combined). Location-specific availability and postal-code-based matching ensure accurate guidance regardless of region.
See it in action
Outcomes
The AI agent handles 85% of customer inquiries across all 29 locations. Package recommendation accuracy sits at 92%, and booking abandonment dropped by 78%. Front desk teams now spend their time on in-clinic service instead of answering the same questions about gestational timing and package details.
The system understands the relationship between gestational ages and service offerings: early pregnancy guidance (7–9 weeks) for gender testing, optimal imaging windows (21–25 weeks) for 3D/4D/5D packages, and late-term support (36+ weeks) for fetal position scans. Customers get accurate, contextual recommendations instead of generic responses.
“The AI agent handles the majority of our customer inquiries across all locations — from package recommendations to booking guidance. Our staff can focus on delivering the in-clinic experience instead of answering repetitive questions. The system understands gestational timing better than we expected, and families are getting to the right package faster.”
— UC Baby Management, Healthcare Consumer Services
What this architecture enables for healthcare networks
The pattern UC Baby uses — tool calling for service matching, knowledge retrieval over a clinical catalog, multi-location support, and gestational-context-aware responses — applies directly to other healthcare and multi-location deployments:
Multi-location clinics
Same tool-calling pattern for service matching and location availability across any number of clinic locations. One knowledge base, consistent responses everywhere.
Patient intake & triage
Context-aware intake that routes patients to the right service based on symptoms, history, and provider availability — reducing wait times and misrouted appointments.
Insurance & billing support
Knowledge retrieval over coverage policies and billing codes. Patients get accurate cost estimates before booking instead of finding out after.
Building something similar?
We can have a working proof-of-concept on your data in 10 days.