Sevalife

How many lives will you save today?

Comment 3

Sevalife

Save lives in 3 steps: commit, confirm, and celebrate!

Problem

Behind numbers, there is life. And with life comes happiness. We need to feel happy for being helpful while we solve the global COVID-19 crisis. Most people do want to help in order to end the current situation faster on a basic level, but they are not aware all the time; even presidents and kings forgot to avoid shaking hands.

However, this effort is helpful. Small actions from you, your friends, and your community can measurably decrease the expected number of infections and deaths in your country. But how can we ensure people are aware?

Solution

Sevalife is a simple chat that helps you take responsibility for solving the Pandemic by performing 3 daily steps:

  1. Commit: asks you in the morning what will you do
  2. Confirm: tells you in the evening how many lives you saved
  3. Celebrate: See how many lives you saved

Sevalife enters the game for letting its users know how beneficial their actions are for today and for the future., and the user gets the following benefits:

  1. A day to day boost on following the recommended precautions during this global Pandemic.
  2. The possibility of doing a recap on the actions and activities done throughout the day.
  3. Notifications in order to keep social responsibility awakened.
  4. A multilingual interface with flows in English, Spanish, French, and Italian.
  5. An estimated number of saved lives (feels good!) using advanced modelling.
  6. Sharing is caring. Boost and encourage friends to keep the Corona fight.

Sevalife is a chatbot implemented in Chatbot.com (chatbot provider in Poland), previously implemented in Chatfuel (chatbot provider in California) and with a glitch.com backend, It runs in Facebook messenger, and it does not require installation.

Progress

Hackathon 3: EU vs Virus

(April 24-26)

This hackathon was meant to help us validate the Chatfuel chatbot so far. We had ambitious goals, but unpredictable events are part of our lives, and we had to sketch an action plan to solve it - as with the global virus.

We were confident to open the chatbot from 2 external users (from last week) to about 10 users (Friday evening). We learned that a chatbot implemented in this platform will run well for a few people, but will fail when as few as 10 people join (from our 1.000 users paid plan).

From the 10 users, only 3 managed to complete the first flow. All others simply got no response back from the chatbot at random steps in the flow. Then the Chatfuel dashboard became unavailable for 30 minutes. We contacted customer support, and before they answered, we logged two more severe issues in their infrastructure.

We decided to fight and change to a new server and, as a result, we ended up prioritising this activity. We have migrated our chatbot to a new platform (Chatbot.com) and have implemented the flows partially, but it will be an ongoing process after the hackathon.

We continued improving on the communications strategy (brand tone, SoMe, copies, and polishing translations), while our developers were migrating the flows (a manual process).

Team: Vlad Manea, Alejandro Trujillo Urién, Vikrant Singh.

Hackathon 2: Global Hack

(April 9-12)

During Global Hack, we improved the chatbot in Chatfuel and Glitch with three features:

  1. opened the interactions in multiple languages: English, Spanish, French, and Italian.
  2. estimated lives saved by using a state of the art, accurate epidemiological model (SEIR), using public data for France, Spain, Italy, the United Kingdom, and the United States; the model estimates the positive impact as the number of lives saved by daily actions of social isolation by a fraction of the population.
  3. allowed the user to invite friends and see a joint leaderboard of estimated lives saved.

Team: Vlad Manea, Emma Barme, Alejandro Trujillo Urién, Florin Tatu.

Hackathon 1: Hack the Crisis Denmark

(April 3-5)

We developed a basic version of our chatbot in Chatfuel and Glitch. The chatbot:

  1. asked questions about daily life, activities, durations, and people met.
  2. helped users commit and confirm simple steps to reduce exposure to getting infected.
  3. estimated the number of lives saved by using a basic, visual model.

Team: Vlad Manea, Sarah MacPhee (corresponding), Morten Sørensen (corresponding), Kate Wac (corresponding).

Impact

Wiseman says Data says "if 10% of your country did your steps, there would be ~40.000 fewer deaths from COVID-19 over the next two months."

Our approach is mainly social and behavioural. We can influence the way our target has access to information sources and positive behaviours for society. It can serve as a portal for sharing good practices and educating our population on topics like hygiene and procedures for “coming back to life”.

Necessities

Our team has an ambitious approach to the future.

  1. We are working on improving the user experience by optimising the flows.
  2. New languages will be added and proofread when our system is stable.
  3. Add new useful features and improve the Activities section.
  4. Collaborate with entities interested in the educational and research potential of the chatbot.

We plan to continue by validating the chatbot with users, first in our countries, as we add more languages. We are now at 10 users we nurture. We have now observed the old platform did not scale, and we are moving it to a more stable platform.

We plan to improve the flows to increase motivation by following a hook model (Nir Eyal): triggers, actions, variable rewards, investments such that people can become more engaged. We can validate these flows and our solution in a case-control study (n <= 1.000 participants).

We plan to obtain help from healthcare and governmental bodies and businesses who can help us build the platform further, as we collect valuable data from anonymous participants.

Set up collaborations with researchers to collect more complex, but easy to obtain, information besides the number of people by using a combination of self-reported and technology-reported activities of daily life to help improve current models estimating social isolation.

Value

Our chatbot can be used by public institutions, official organisations and private Healthcare companies to make use of this way of communication and population-level research with citizens. The idea was made for lasting in time, as it will help in the future with other pandemics and sicknesses.

Sevalife gives easy access to guidelines on how to handle a global health crisis, how to be aware, and what is the estimated positive impact of doing so. Sevalife provides accurate information, ready to be customised to the person’s location.

Our team thinks globally, and we want this to be helpful for a lot of people over the world across national, cultural, and technological barriers. We use a simple platform (such as Messenger) for a complex purpose, to keep uncertainty with “too much” technology at bay.

While becoming a secure data source of information for research and society, we imagine Sevalife collaborating in the future with authorities and researchers deriving policies against our microscopic enemy. For example, using activities of daily life between crises to estimate baseline social exposure, and activities of daily life during crises to estimate social isolation (e.g, studies on N >= 1.000 participants).

Built With

Share this project:

Updates