Skip to content

CompanionCare Design science research at McMaster University

Connected care.
Built around people.

An AI companion helping older adults stay connected to their care team, from the comfort of home.

Architectural pencil sketch of an older adult at home and a clinician in a separate care workspace, joined by fine communication lines.

A connection between conversations.

The AI companion brings conversation, context and follow-up together, with privacy built into the system.

Explore the system
~250 ms

Reported prototype response

Open by design

Self-hostable technology

2 papers

Peer-reviewed research

Technology that makes
room for people.

Multimodal chat

A conversation at your own pace.

Talk naturally, pause when you need to, or switch to text. The companion adapts to the person using it.

Lifelike avatar as the primary interface
Low-latency voice conversation with interruption handling
Local end-of-turn model plus adaptive pacing and tone
Technical notes

Older adults interact through voice or text in the Care Companion app. An low-latency agentic AI stack (project-reported prototype average of about 250 ms) handles each turn with natural overlap: adaptive interruption handling, an on-device end-of-turn model so the agent does not cut in too early, and adaptive voice speed for clarity.

How it works

A closer look at the connection.

How the voice-and-video care companion connects patient experience, intelligent workflows, and secure provider tooling in one line of sight.

From everyday conversation to context a care team can use.

Select a stage to explore

Real-time agent

The agent coordinates conversation, responds with voice and an avatar, and handles natural interruptions without losing context.

View the original research diagram
System architecture
Technical diagram of multimodal input, the real-time AI agent, care workflows, secure memory and provider insights.

Technical diagram of multimodal input, the real-time AI agent, care workflows, secure memory and provider insights.

Designed for consent, oversight and accessibility

Consent, traceable AI outputs and encrypted connections guide the privacy design. Live transcription supports people when hearing is limited. Each deployment needs its own security and governance review.

Privacy & consent

Dual patient consent, encrypted transit and at-rest data, audit logging, and no sale of conversation data to third parties.

Open & self-hostable stack

Components and models are composed from open-source, self-hostable software so PHI and workloads can stay on infrastructure you control—not opaque SaaS you cannot audit.

Transparency for clinicians

Dashboard summaries link back to source messages so providers can verify AI-generated text.

Human in the loop

The chatbot does not diagnose; escalations and provider review keep decisions with care teams.

Built for natural, private conversations

Voice and avatar tuned for clarity, with guardrails and encryption so sessions stay in environments you control, on open, self-hostable services suited to regulated care.

Adaptive interruption handling

Smooth barge-in and overlap so users are not locked into rigid turn-taking when they think aloud or self-correct.

Lifelike avatar interface

A visual companion accompanies voice and text. The research evaluates how this interface supports clarity and engagement.

Low-latency voice agent

The project reports an average prototype response of about 250 ms. Actual performance varies with the model, hardware, network and deployment configuration.

Intelligent end-of-turn

On-device end-of-turn detection cuts down on clipped sentences and premature agent talk-overs.

Adaptive voice speed

Playback pacing adjusts for clarity—slower when needed, without sounding robotic or rushed.

Secure & private connection

Encrypted sessions with no third-party data sharing for unrelated analytics, ads, or model training—aligned with stacks you can self-host when policy requires it.

Publications

Built, tested and shared with the community

CompanionCare is design science research: we build the system, put it in front of older adults and care providers, and refine it with what we learn. Publications and community presentations document the project as it develops.

Read the research overview and evidence

Peer-reviewed publications

The people behind the research

A doctoral design science research project at the DeGroote School of Business, guided by faculty in information systems and computer science at McMaster University and East Carolina University.

  • Portrait of Emad Deilam Salehi

    Emad Deilam Salehi

    PhD Candidate, DeGroote School of Business, McMaster University

    Contact card
  • Portrait of Dr. Yufei Yuan

    Dr. Yufei Yuan

    Professor of Information Systems

    DeGroote School of Business, McMaster University

  • Portrait of Dr. Kamran Sartipi

    Dr. Kamran Sartipi

    Associate Professor of Computer Science

    East Carolina University · Adjunct Associate Professor, Information Systems, McMaster University

  • Portrait of Dr. Shane Saunderson

    Dr. Shane Saunderson

    Assistant Professor of Information Systems

    DeGroote School of Business, McMaster University

FAQ

Common questions

Something else on your mind? Email deilamse@mcmaster.ca.

Research overview and sources →

What is CompanionCare?

CompanionCare is a design science research project at the DeGroote School of Business, McMaster University, building an AI voice companion for older adults living at home. Older adults talk to a realistic avatar by voice or text, and their care team follows along through a secure clinician dashboard that summarizes conversations and flags what needs attention.

Why is CompanionCare built as design science research?

Design science research builds a working artifact to solve a real-world problem, evaluates it with the people it is meant for, and improves it over repeated build-and-evaluate cycles. In CompanionCare, the artifact is the voice companion and its clinician dashboard. Each version is shared with older adults and care providers through demos and poster presentations, including the MIRA & Labarge Knowledge Exchange, Mac engAGE and the AGE-WELL Annual Conference, and their feedback shapes the next iteration.

Who is CompanionCare for?

CompanionCare is designed for older adults aging at home, their caregivers, and the clinicians and care organizations who support them, such as clinics, home-care agencies and senior-living communities.

Does CompanionCare replace doctors or make diagnoses?

No. CompanionCare is a research prototype and does not diagnose or replace a healthcare professional. The design includes escalation workflows and provider review of AI-generated summaries linked to the original messages. Medical decisions stay with the care team.

How does CompanionCare protect privacy?

CompanionCare uses open, self-hostable components so a care organization can control where health data is processed. The privacy design includes consent, encryption, access controls and audit records. Each deployment needs its own security assessment, data governance and institutional approvals; this website does not establish regulatory compliance.

How natural is the voice conversation?

The project reports an average prototype voice response of approximately 250 milliseconds; actual response time varies with the model, hardware, network and deployment. The prototype supports interruptions, on-device end-of-turn detection and adjustable speaking pace. The figure is not a clinical outcome or a service guarantee.

Who leads and supports CompanionCare?

CompanionCare is led by Emad Deilam Salehi, a PhD candidate at the DeGroote School of Business, McMaster University. The research team includes Dr. Yufei Yuan, Dr. Kamran Sartipi and Dr. Shane Saunderson. Support comes through the 2026 MIRA & AGE-WELL EPIC-AT Fellowship. The research overview links to the institutional project record and original publications.

How can I take part in a study or partner with the project?

Older adults and caregivers can join guided sessions of 60 to 90 minutes, held remotely or on campus. Clinics, home-care agencies and senior-living communities can partner on pilots, co-design and evaluation. Email deilamse@mcmaster.ca to get involved.

An open invitation

Better care starts
with your perspective.

Help shape technology for older adults. We welcome participants, caregivers and clinical partners.

Take part in a study