From my experiences in social work, I know that immigrants often face difficulties in contact with the health care system. That’s why I decided to create a service that would bridge cultural differences in health communication.
Use Interviews, Desk Research, Workshops, Chatbot Scripting, Wireframes, Prototyping, User Testing

The user might be misunderstood when communicating with health care professionals.

Health conditions can be expressed very differently in different languages and many cultures use a much more poetic and metaphorical language to express illness than in Swedish or English.

The user might not know Swedish or English well enough to be able to express their health issues clearly enough to get the help they need.

Mental health issues are very stigmatized in many cultures and the user might not seek help for it.
Metaphorical ways of expressing health conditions in different languages.

The world is taking me left and right

They were hunting me with horses all night

My head is boiling like fire

AI based chatbot / voicebot including an interpretation service that takes cultural context into consideration and is trained to recognise methaphorical health expressions from different languages.

Health advisor with AI trained to notice a combined tracking of several of the physical symptoms of depression/mental health issues like low energy levels, headache, stomach pain, chest pain, insomnia, changing appetite, restlessness etc. and help target group access accurate health care in a sensible way.



The final prototype is built in Protopie based on exported Figma wireframes, and features voice interactions.
Here’s a link to the prototype.
The voice function only works in the Protopie player though, since it’s a beta function.

Updates based on feedback
from User Testing:

Added the option to translate text to native language.

Added the option to edit message to send to GP.

During my work with this project, I reflected on ethical issues and challenges that might arise if this was to become a real product.

Voicebots are biased towards mainstream accents and would need a lot of training to understand the accents of newly arrived immigrants.

Gathering a big enough database of health expressions in different languages and having health care professionals to script the chatbot properly would be very time consuming and require a lot of resources.