There are many challenges that people that isolate at home face. Whether self-isolating or being quarantined, there are is a wide range of needs that are more difficult to meet: doing your shopping (from essentials to small nice things that make isolation more bearable), reaching out to your doctor (or any doctor) for problems or simply questions, having someone to talk to and so on.

Fortunately there are people willing to help. How to reach those that are in need?

Inspiration came when I saw a couple of flyers left around the neighbourhood by a couple of NGOs, each saying something to the extent of "are you (or do you know) someone who is isolated or in the groups of high risk at contracting the virus? Contact us so that we can help you be less isolated. We can help with your shopping needs ... etc".

There must be a better way.

The problem

People in isolation face a number of challenges. Getting their daily supplies, getting access to medical advice (for themselves or for their pet), finding someone to share their daily struggle with, someone willing and able to take their pet for a walk.

While there are services out there that provide some assistance with a subset of these needs (for example, there are delivery services such as Glovo, Uber or Bringo - but they focus on urban areas and typically they don't focus on or prioritize service for those particularly at risk), they either don't cater to those that are particularly at risk or don't have enough coverage.

Proposed solution

Care@Home is a solution meant to be delivered primarily as a mobile application (ideally as a device+application, for those most in need that may not have a suitable device) aimed at allowing users (beneficiaries) to express their needs. These needs can be categorised (practical -> shopping, medical -> advices or checkups, social -> get in touch with a psychologist or simply someone available for a talk/chat). Once defined, a need (or task) is dispatched to a number of local volunteers or organisations (providers) that have signed up as available to satisfy one or more types of needs.

The application does not aim to fully manage the beneficiary / provider relationship, but it aims to establish relationships while also providing basic contact tools (chat, video calls, notifications, log of requests) all the while aiming to be straightforward to use. Given the target audience, the application's interface aims to be primarily based on natural language (spoken commands) with a classic UI as a fallback.

What makes Care@Home different

Extendable system of needs

Initially we are focusing on three types of needs: medical (where vetted volunteering medical professionals offer their time to offer medical advice or professional psychological support), shopping (where providers can take up and fulfill shopping lists or medical receipts, support (various tasks like offering emotional support through conversation or dog-walking). However, the system can be extended as each need consists of a flow of actions that takes a beneficiary through defining what they need and on the other hand the provider being able to subscribe to be notified of tasks for that flow.


All participants, whether they are beneficiaries or providers, will be validated. If a doctor volunteers their time to provide medical advice or checkups, their credentials should be validated. On the other side, since the app is aimed at those at risk, the situation of the beneficiary should be verified as well.

Speech-based interaction

To make it easier for people less familiar with mobile devices to use the device, interactions are intended to be primarily done through natural language which is parsed down to keywords and commands resulting in actions. To support those that are unable to interact via speech and sound (or aren't comfortable with that), Care@Home can fall back to a more classical UI.

Simplified workflow

As a volunteer (or organisation), you choose what types of needs you want to help with, then for each need choose a geographical area in which you want to provide it.

As a client (beneficiary), you tells us about yourself as much as you want us to know (though things like name and location are considered a prerequisite)

Define whom you want to help, when and where

As a volunteer you may want to define:

  • what needs you can fulfill and where (for example, you may be able to help the elderly on your street, but still you can talk to and offer your time for emotional support to anyone in the country)
  • whom you want to help (perhaps you want to specifically prioritize those disabled and have their tasks reach you with priority, or perhaps you want to help those essential workers working in hospitals)
  • between what hours you are available (you may want to take shopping tasks between opening hours of your closest shops, but for emotional support you can be reached at any time)

Leverage existing logistics or knowledge

Care@Home doesn't handle the underlying logistical challenge but instead enables individual volunteers or organizations to efficiently use their resources and ensure that they don't overlap.

Perhaps an organization targets the elderly in isolated rural areas and simply want to be able to establish clear communication and supply lines while also enabling those individuals to seek out emotional support.

Perhaps an organization is already helping disabled people in particular and wants to have a better reach.

Will it help beyond the crisis?

The novel coronavirus has pushed more people into isolation but unfortunately there were people and communities in various degrees of isolation before the current crisis and there will be people in need of help after this as well. This application is meant to facilitate access to helpful services and we believe there will be a need for it later on as well.

How we built it

Our aim is to offer a mobile application that can be bundled with specific devices so that organizations can offer it as-is to those that may not have the means to acquire mobile devices on their own. At the same time, the mobile app is meant to be available so that it can be installed on any Android device.

In addition, the focus on speech interaction should allow us to provide a version of the application that can be delivered through digital-assistant enabled devices (such as small speakers) without the need of a tablet.

Challenges we ran into

As a technical team, an important challenge that we found was to nail the scope of what we wanted to achieve during the hackathon and not be distracted by technical goodies discovered along the way.

Another challenge was defining a product that caters to the needs of potentially less technical people than us. Being people who grew up surrounded by technology, shifting away from this mindset into something completely different is a tremendous undertaking.

A technical challenge that we faced (and haven't yet solved) is to create a better voice interaction component that would better fit the profile of the application (one that, for example, can recognize the spoken language as the user speaks and that can be reliable through different languages and accents).

Accomplishments that we're proud of

We've managed to put together a working application that makes use of speech interactions for a basic flow: registration, shopping list definition and issuing the task.

What we learned

  • we need to get better at pitching
  • how to adjust technical solutions to cater to potentially less technical people than us
  • Google Cloud has such amazing tools for developers!

What's next for Care@Home

In addition to implementing the proper flow for the rest of the needs, we want to incorporate some machine learning techniques to help sort tasks and distinguish items in order to make the lives of the providers easier.

We want to approach organizations to try out our solution (we already approached some that seemed interested) together with people they already help.


The attached APKs represent the two applications: one for the home users (beneficiaries) and one for providers. While they are functional, the phone validation via SMS code works for phone numbers registered under our demo Twilio account only.

Share this project: