Research at McMaster University
AI companions for older adults.
The research behind CompanionCare.
CompanionCare is a design science research project at the DeGroote School of Business, McMaster University. It explores how an AI voice-and-avatar companion and a clinician dashboard can support communication between older adults aging at home and their care teams.
Project overview by Emad Deilam SalehiUpdated
Read the publicationsProject at a glance
- Research focus
- AI-supported communication between older adults at home and their care teams
- Approach
- Design science research: build, evaluate and refine a working prototype
- Institutional home
- DeGroote School of Business, McMaster University · Hamilton, Ontario, Canada
- Lead researcher
- Emad Deilam Salehi · PhD Candidate
- Support
- 2026 MIRA & AGE-WELL EPIC-AT Fellowship
- Project status
- Research prototype; not a replacement for clinical care
What is the project investigating?
The project investigates communication, accessibility and human oversight in AI-supported home care. The research method is iterative: build a working system, gather feedback from older adults and care providers, and use that feedback to refine the next version.
A companion for everyday conversation
Older adults interact by voice or text, with an avatar as a visual interface. The prototype supports interruptions, turn detection and adjustable speaking pace. These features are intended to make conversation easier to follow; their usefulness is a subject of evaluation.
A dashboard for the care team
A separate clinician workspace brings together conversation summaries, topics and follow-up information. Links to source messages allow providers to check AI-generated summaries. The design keeps clinical review and medical decisions with people.
Privacy and institutional control
The architecture uses open, self-hostable components to support deployment on infrastructure a care organization controls. Consent, access controls, encrypted connections and audit records are part of the privacy design. A deployment still needs its own security assessment, data governance and applicable institutional approvals.
Explore the interactive system diagramPublished work
Publications and primary sources
Use the original publications when citing the technical or ethical research.
Generative-AI Solutions for Connecting Seniors and Healthcare Providers
Emad Deilam Salehi, Yufei Yuan, Kamran Sartipi
2025 IEEE 38th International Symposium on Computer-Based Medical Systems (CBMS) · 2025
DOI: 10.1109/CBMS65348.2025.00092System design research on a multimodal companion and an AI-supported provider dashboard.
In Whose Hands? Ethical Judgment of Generative AI in Senior Home Care
Emad Deilam Salehi, Yufei Yuan
Proceedings of the 2025 Pre-ICIS SIGDSA Symposium · 2025
Research on ethical judgment and responsibility in the use of generative AI for senior home care.
Project and fellowship record at MIRA
AI That Listens: Co-Designing a Safe, Multimodal Communication Companion for Older Adults Aging at Home. Institutional project record from the McMaster Institute for Research on Aging.
How should the research be interpreted?
CompanionCare is a research prototype. Publications describe the research questions and system design; publication alone does not establish clinical effectiveness, regulatory approval or suitability for a particular patient. The companion does not diagnose or replace a healthcare professional.
The project reports an average prototype voice response of approximately 250 milliseconds. Response time depends on the model, hardware, network and deployment configuration. This is a prototype performance figure, not a clinical outcome or a service guarantee.
For study protocols, evaluation details or collaboration enquiries, contact the research team rather than treating a demonstration as a clinical service.
Contact the research teamWho is behind CompanionCare?
The project is led by Emad Deilam Salehi, a PhD candidate at the DeGroote School of Business, McMaster University.
- Dr. Yufei Yuan
Professor of Information Systems · DeGroote School of Business, McMaster University
- Dr. Kamran Sartipi
Associate Professor of Computer Science · East Carolina University · Adjunct Associate Professor, Information Systems, McMaster University
- Dr. Shane Saunderson
Assistant Professor of Information Systems · DeGroote School of Business, McMaster University
Support comes through the 2026 MIRA & AGE-WELL EPIC-AT Fellowship.