Boston Consulting Group / New York / 2018-19
Rethinking elderly care using AI.
Co-founded elovee, a healthcare startup focused on developing technology to improve day-to-day care for seniors with dementia. It was the first virtual-care AI-enabled voice and video platform for seniors living with cognitive decline, that helps caregivers better address behavioral and emotional needs of seniors at care homes.
Introducing elovee
A personalised dementia care platform that combines the familiarity of family relationships with the expertise of dementia experts.
With elovee, caregivers at care homes can use recorded videos and digital renders of family members to reassure seniors when they are refusing care, and when family members can’t be present themselves.
Context
Incubated elovee in 2019
In 2019, while I was working with BCG’s digital arm Digital Ventures, my team and I were tasked with partnering with a chain of luxury care homes in Canada to revolutionize elderly care.
Over a period of a year, our cross-functional team focused on ideating, validating, incubating and launching a company called elovee, that raised $13.5M funding right after the launch, and $30M total.
Role
Research and design (team of 3)
I was a part of the 0 to 1 founding team that created elovee and one of the three designers that worked on the product from ideation to launch.
I was leading research, synthesis and product market fit efforts at first. Later, I was responsible for designing the caregiver experience through the whole design process including feature exploration, prototyping, testing, iterating, etc
Initial process and insights
-
Caregivers need more effective tools to reassure seniors
Most care homes, caregivers needed easier and more effective ways to make seniors receptive to care. Caregivers were often taking care of 5 patients at a time which made them encounter a lot of high stress situations on a daily basis.
-
Seniors need specialised and individualised care to stay calm
There were four major behaviour patterns that seniors experienced - agitation, angst, fixation and boredom. Each situation requiring specialised and individualised immediate care to reassure and calm them down, often in the middle of the night.
-
Family members often feel guilty and unsure of how to help
Family members often carried a lot of guilt for not being able to spend more time at the care home, as often life responsibilities came in the way. They also felt detached from their loved ones and expressed a need to have more constant updates on their wellbeing.
Design Principles
❤️
Seniors with dementia
Provide a series of activities and experiences that help them feel relaxed and connected.
👨👨👧👧
Family members
Provide a day to day companionship, reassurance and education platform, that helps ensure wellbeing of their loved ones and gives them peach of mind.
⚕️
Caregivers
Provide personalized care, data trends, and tools to communicate with families and each other that help them do their job more efficiently, increasing the number of positive interactions.
💼
Executives
Provide a cost effective tool that reduces their cost of care and increases their customer satisfaction.
Approaching the problem
-
Diverge: Identifying opportunities and generating ideas
We explored different interaction models, technologies and mediums to deliver care. Our solutions ranged from creating a pet robot companion to keep a senior engaged on a daily basis to a video/voice assitant that kept them in touch with their family members at all times.
-
Converge: Concept and value proposition testing
To gain confidence in a direction, we conducted interviews (storyboards, prototypes and card sorting) with caregivers and family members. We learnt that interactions with family members, in person or even on a call, was sometimes we most effective way to calm seniors down in times of need.
Seniors with dementia often thrive in the presence of familiar faces, particularly those of family members. However, consistent family presence isn't always feasible.
In such cases, amalgamating the warmth of a family member's presence with the specialized care of a dementia care expert can significantly enhance their overall well-being.
Incubating the concept over a year
-
Building the conversational model for reassurance
We worked with an expert in cognitive wellbeing in the dementia space, to design a conversational user interface (CUI) that followed a four step conversational model: seek to understand, validate, reassure and redirect.
-
Video and voice replication for reassurance calls
I partnered with a team of data engineers, Unity developers and product managers to deliver a quality experience for seniors. My main role in this stream was to make sure the visual and voice representation of the avatar was in sync with natural conversation.
Caregiver experience
Enable caregivers to effectively reassure seniors
I spent the most of the project honing the caregiver experience of providing care to the seniors. I worked with a PM and engineers in two week sprints to define, design, test and build an app for caregivers.
Interestingly, this RCP ‘app’ was completely designed in Unity which brought with it added complexity for design and engineering.
Principle 1:
Designing for busy and overworked caregivers
Principle 2:
Asking for feedback to keep improving care
Principle 3:
Keeping shared knowledge between different caregivers
Pilot study with 17 family members and 15 residents
“She is usually a sad, worrying person. We were really struggling but after adding elovee to her care plan, she had started smiling more.”
“I’m so excited to be able to help build this platform – seeing mom feel calmer and more cheerful these days, it’s means the world to me, just wonderful.”
“I see this platform as a great way to address our labor crisis, which has been going on for years. It could decrease turnover and increase efficiency.”
“I would definitely use it. It really calmed Lolly down today. Seeing Syd changes her mood, you saw it – she was kissing the screen.”
“I think it is absolutely amazing. It warms my heart to be able to help mom, especially when I’m away for work or taking care of children”
“On the last day, two seniors were sitting down chatting with each other, the RCPs had never seen that before and both were pilot participants”
Synthesized learnings to six major
ethical AI principles
🎛️
Advocate Autonomy
Are we giving people the capacity to self regulate and control this service throughout the journey?
🟰
Ensure Equality
Are we listening to the needs for all our stakeholders? And are we solving for those needs equally?
💑
Cultivate Connections
Does our solution detriment personal relationships or reduce meaningful interactions among our stakeholders?
🪟
Target Transparency
Are we holding any information from the stakeholders that might influence their decision to use this product/service?
🧛
Integrate Inclusivity
Are we enabling people with diverse backgrounds to use our product in different environments?
💆♀️
Warrant Wellbeing
Will this ever be harmful to an individual’s physical or emotional state or wellbeing?
Creating Alignment via Branding
Outcome
Raised a total of $30 million
Got 10 other residences centres to sign up for a pilot.
Raised a total of $30 million over two rounds of investment.
Got featured in The Economist as a potential gamechanger in elderly care.
Currently, elovee has pivoted to a more focused product that is using telephonic calls rather than video to provide care.
Key Learnings
Designing for AI requires setting strong ethical principles
Importance of incorporating design ethics in our everyday work. Post this project, I started creating my own ethics methodology.
Owning design throughout the design cycle from research to visual design
Designing for multiple user groups for multiple use cases
Using iterative testing with multiple user groups to improve product. More importantly, involving the whole team in research and testing.
Technology/Products can’t solve for all the problems and so mitigating its risks is important