Deployment of a Multilingual, Embodied, Holographic Artificial Intelligence Concierge in Ambulatory Care: Early Implementation Lessons From the Artificial Intelligence Responsive Interactive Agent
Year: 2026
Type: article
Abstract: Health care environments can be difficult to navigate, particularly for individuals with limited language proficiency, hearing impairment, low digital literacy, or unfamiliarity with the facility. Conversational artificial intelligence (AI) systems may support accessibility and navigation, although descriptions of real-world implementation in health care settings remain limited. As a descriptive, single-site quality-improvement pilot, this report characterizes the experience of implementing a conversational AI agent in the lobby of a large, outpatient, academic practice. The Artificial Intelligence Responsive Interactive Agent (ARIA) is a multilingual, 3-dimensional, holographic conversational AI concierge developed to provide real-time assistance with navigation, digital onboarding, and information regarding amenities and services. ARIA leverages natural language processing and large language model technology to provide voice-enabled interactions in English, Spanish, Mandarin, and Arabic, with real-time closed captioning. The system can also generate quick response codes that push information directly to users' smartphone devices. Piloted in a large ambulatory lobby from October 21, 2025, to February 2, 2026, ARIA recorded 1429 device-activation sessions. Of these, 689 sessions (48%) contained analyzable user interaction data. Wayfinding and general information requests were the most common use cases. Deployment surfaced operational challenges, including audio capture, network reliability, interaction timing, workflow integration, quick response transfer, and content governance. As the use of AI in clinical settings expands, there may be a role for an embodied agent serving as a digital concierge if these barriers can be addressed. Future evaluations should prospectively define implementation outcomes, use systematic utilization denominators, and assess user experience and operational impact with validated measures.
Authors: Ellen C. Meltzer, Jared T. Staal, Alyssa B. Chapital, Pete J. Pallagi, Katelyn M. Cheetham, Haley Wang, Qiuyu Jin, Rafael Fonseca, Aman Anand, Mark R. Stemler, Chris S. Jaeckel, Danielle M. Stricker, Jacob M. Dison, Cherish D. Caputo, Tharana Maran, Beverly Hansen, Praneetha R. Elugunti, Biju T. Samkutty, Anjali Bhagra
Institutions: Mayo Clinic, Mayo Clinic in Arizona, Mayo Clinic Hospital
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $2,500
Funder: Mayo Clinic
OpenAlex ID: W7221086638