top of page
image.png

Mindfulness Conversational Agent

User experience Research and Product development for persons with dementia and caregivers

Problem

​Dementia is one of the most prevalent global public health challenges. 78 million people are estimated to be affected & US $2.8 trillion dollars cost by 2030 (WHO, 2021). It is a distressing and challenging disease that causes high levels of stress and worries for both persons with dementia and caregivers, especially as they continuously decline. Promising mindfulness interventions can address these issues and provide the pair with coping mechanisms.

Mindfulness cultivates present moment awareness, and the development of a nonjudgmental lens on experiences, which may be particularly useful for helping persons with dementia and their caregivers to not dwell on the past or future, which causes worries and anxiety. It also encourages learning to respond instead of react, which facilitates thoughtful responses to experiences.

Studies show promising preliminary results for paired approaches where persons with dementia and caregivers practice mindfulness together. Conducting paired mindfulness using conversational agents like Alexa is novel, and it is able to provide high levels of accessibility and scalability, while providing more guidance.

This novel intervention was developed during the Covid-19 crisis, when social isolation was particularly challenging. As dementia condition declines, persons with dementia and carers find it increasingly complex to meet needs, causing stress, worry and a loss of perceived control which strains relationships. The pandamic, at its’ peak, made physical meetings dangerous and seniors without digital savviness found it hard to stay connected. While mindfulness practices offered benefits, they were often difficult to implement given that they were often hours long and done in person with experts. Voice-based technologies, such as conversational agents, was a more accessible alternative, yet few interventions were specifically designed to support both individuals as a connected caregiving pair.

Question: How can mindfulness activities on Alexa be designed to help persons with dementia and caregivers?

1> User interviews

2> User centered Design prototyping

3> Feasibility study

Process

In depth user and expert semi-structures interviews were done based on a topic guide through a 1-hour interview.

 

Interviewees include dementia experts, mindfulness experts, persons with dementia and caregivers

4 iterative user centered design cycles with prototyping was done, improving on each prototype at each stage, before a working product was created.

 

Interviews, workshops, ethnographic research was done. Each prototype was presented to stakeholders for feedback and changes were made based on the feedback.

30 day feasibility study with a working product to evaluate effectiveness.

 

Assessments, dairy records, surveys and interviews done before and after activity usage revealed positive results.

User interview Insights

Themes from the interviews were extracted from transcriptions through a thematic analysis approach on Nvivo. They were also compared to existing products and research to determine novelty and effectiveness.

These themes were used to create the first prototype to generate user feedback.

 

Regarding preferences, mindfulness was likely to be used by all interviewees using a paired approach through conversational agents.

Screenshot 2025-08-05 at 18.38.15.png

User Centered Design Prototyping

Through a user centered design approach involving a widespectrum of stakeholders, I created Alexa skills to guide the pair through mindfulness activities. There are 4 activities available, and the first one encourages them to stay in the present moment through mindful breathing. Other activities like making space, focuses on metacognition and cognitive defusion where it encourages them to think of a stream and visualise any and all of their thoughts on leaves, floating away, without engaging in the thought. This helps separate themselves from their thoughts, viewing them as transient experiences. There are also gratitude and kindness exercises.

Four design iterations of each activity resulted in a working product and an understanding of user needs. The product was accessible and appropriately timed, with users expressing a preference for regular use that could easily integrate into their daily routines. Findings indicate that it has strong potential to support mindfulness practices for the pair.

 

In addition to the product, both broad and nuanced design guidelines were identified. These included detailed descriptions of each need, along with practical examples for how the guidelines could be applied in future designs.

Screenshot 2025-08-05 at 18.45.30.png

30 days Feasibility study

10-15 mins activities



2-2.5 hours long

Done daily




Weekly sessions

Embedded into daily life



Face to face formal sessions
Fixed physical location
Fixed time
Fixed completion period (8wk-1Yr)

Facilitates choices



Structured

Designed for pairs



Not personalised

Results

Successful Results

Reduced Stress (83% of participants)
Reduced Worry (75% of participants)
Greater sense of control (83% of participants)
Better perceived closeness (41% of participants)

Special consideration was given to accessibility needs, addressing cognitive capabilities, language simplicity, speech speed, tone, pauses, and overall therapeutic impact.

Participants liked the intervention and found it accessibly and engaging, fitting into their lives seamlessly. Activities were also likely to be done regularly after the study ended.

Screenshot 2025-08-05 at 19.03.09.png
User Personas Generated
Screenshot 2025-08-05 at 19.15.55.png

Personas were created based on user data and feedback for future designers to utilise

Screenshot 2025-08-05 at 19.16.06.png
Screenshot 2025-08-05 at 19.16.18.png
Screenshot 2025-08-05 at 19.17.12.png
Screenshot 2025-08-05 at 19.17.27.png
Screenshot 2025-08-05 at 19.17.35.png
bottom of page