#AIAccelerator2020 Winners: Eleos — Changing the way young people manage their mental health.
We are Eleos, a group of four teens — Anisah (Project Lead & Manager), Imani (Design Thinker), Mohammed (AI/ML Developer) and Israel (Researcher) — who are changing the way young people access support for their mental health and emotional wellbeing.
Eleos started with one call; all four of us, from different ages, backgrounds and schools based in London and Yorkshire. Although none of us knew each other in-person or online, it did not take long for everyone to open up and discuss their interests and ideas of what kinds of challenges they would like to solve. It was at this point that we realised that we were all passionate about mental health and motivated to help support young people with their emotional well being. Put simply, we had the consensus that young people today can be easily overwhelmed with the breadth of resources and support available out there to manage their mental health, whilst being underwhelmed with the unspecific and generic advice that fails to meet their emotional needs. We wanted to change this and through hard work, commitment and the support from our mentors, we spent the next two weeks, dedicating ourselves to create a solution to this problem.
Our hard work, focus and dedication in creating an emotionally supportive platform that helps young people manage their mental health birthed our product, Eleos. Eleos is an AI-powered assistant that aims to curate and deliver targeted advice and external support resources designed to help young people manage their emotional well being needs. We achieved this by creating a unique and personalised dashboard that is tailored to the individual user based on the answers that they give to Eleos. This dashboard offers general advice to learn about your health, specific advice based on the key emotional influencers of each user, self care advice based on user interests to provide long-term emotional support and resources for specific life experiences such as bereavement or bullying.
From the get go, we understood the importance of making young people the center of our project and adopted a process called ‘human-centred design thinking’ that allowed us to solve the complex problems surrounding young people’s mental health in a highly user-centric manner. We began our project by investigating our problem through primary research with young people using a qualitative survey. Our survey garnered 71 immensely detailed responses that allowed us to identify the pain points that young people experience in relation to managing their mental health. Our survey found that 67.5% of young people need support with their mental health but don’t know where to find it, with 87% having gone 3 months without accessing any form of support. Using the user insights we gathered from our research, we built several personas of archetype users that would allow us to design a product that addresses the different emotional wellbeing needs of different young people.
Our Personas laid out on Google Jamboard.
To supplement our primary research, we also explored a number of resources from mental health charities and academics with the aim of getting a better understanding of the underlying issues of our problem. It was at this point that we then began working with our NHS mentors from the Tavistock and Portman Trust. Their expertise in helping young people manage their mental health was integral in shaping the way we designed our user experience and modelled the user journey, ensuring we avoid diagnosing young people and instead focus on identifying their needs and providing appropriate support.
After days of product ideation with support from our design thinking mentors, we began creating a frontend prototype using our design tool, Figma. Through numerous iterations and redesigns, we improved our prototype to address the feedback we received from young people through interviews and surveys. View and test our prototype.
Once we had gained a secure understanding of what young people would want from our product and how best to tackle the issue of excess unspecific advice on the web, we set out to create a backend prototype: the brains of our product. With the guidance from our technology mentors at Mastercard, we developed a machine-learning powered recommendation engine in python that is able to use keywords to accurately recommend to users, advice, apps and websites that meet their preferences; made possible through a technique known as ‘word vectorisation’. In addition to recommending resources based on an individual’s likes and dislikes, we also leverage feedback from users who share similar interests to make recommendations that are inspired by what young people have actually found useful. These two systems of recommendation are known as ‘content-based and collaborative filtering’ respectively and are central to the way in which we meet young peoples’ need for specific resources tailored to them. To aid this process of gathering information about a user’s needs, we also employ a keyword extraction API that is able to scan text input and identify key emotional influencers such as school, exams or family life, before making suitable suggestions of next steps and forms of support.
In creating our frontend prototype and backend model, we worked closely with our Ethics mentor in order to identify potential ethical concerns and make changes to our product accordingly. Although there are still concerns we would like to further address, such as data protection and storage of sensitive user data, we have done our best in ensuring that our product uses emphatic language to create a safe, trustworthy environment for young people and avoids implying any form of medical diagnosis.
Having an idea is one thing, but having a legitimate completed business plan is another story. With support from business mentors, we created a business plan. Our business model included Eleos’s routes to market, revenue streams and the cost structure for our product. Our routes to market include partnering up with the NHS to work on their children’s website for mental health, schools, utilising social media and producing leaflets. Revenue streams include licensing fees and appropriate listing fees. Cost structure thought about the fees relating to marketing expenditure i.e. advertising, hosting and product development.
Pitching our product together live in front of a panel of judges and a large audience was a thrilling experience. It was an opportunity for each of us to showcase our passion for mental health and drive to help young people manage their emotional wellbeing. This was demonstrated not only by our enthusiasm in presenting our product Eleos, but also in our responses to the questions posed by the judges. Although we weren’t very nervous then, the pressure we felt waiting for the results of the Accelerator was beyond nerve-wracking. This made achieving first place even sweeter as we each felt the past two weeks of hard work and commitment pay off, with lots of excitement and screams of joy. Watch our Pitch at the AI Accelerator 2020 Demo Day.
However, our journey does not end here. Our passion for supporting young people and their mental health is why we wish to take Eleos beyond the AI Accelerator and to the individuals we seek to help. By working with the NHS, medical professionals and industry experts, we hope to turn Eleos into a reality.
A huge thank you to Elena for giving us all this opportunity and to all of our mentors for providing us with their continued support throughout the AI Accelerator 2020.
Eleos — made for young people by young people.
For more information about the AI Accelerator: https://www.teensinai.com/accelerator/