The Covid-19 pandemic has spread around the world not only a virus, but as a global economic crisis. The virus transmission coefficient is high (1 person can infect other 40), which has caused many establishments such as shops, restaurants and sports facilities, to close or face strict regulations in order to keep social distancing when operating. In total, the economic effect of the pandemic has brought a 3.5% increase in global unemployment with respect to 2019 (IMF). In Colombia the unemployment rate increased to 21.4% One of the main factors is found on the restrictions on in person transactions (restaurants, stores and so on). Developing countries like Colombia cannot keep the economic at a standstill, for the social consequences are made more critical by the limited aid that government is able to provide. In face of this restrictions, owners had to close their business and fire employees because of the losses. Because of the lack of strategies of safely not only open establishments, but also re-activate economy, we created Kairos, an intelligent crowd-managing platform to decrease establishments closing rate.

What it does

We built a mobile app that lets the user make reservations, know the capacity of the store at a specific time, and predicts the people that will be in that establishment at that hour. This makes for a safer, more efficient visit. Being a two-front platform, the owner (admin) will also benefit from this, having a better sense of the potential visits during and after the lockdown, which not only optimizes the flow of people and hence the labor costs, but also protects and tracks the health of their visitants.

How we built it

In order to train the Machine Learning model we took into count 9 different variables that directly affects people’s mobility, and the one that will be tuned which is the maximum number of people that a place can hold, according to social distancing regulations and its area. All these variables together show a better, but not still exact, way to know the amount of people presented in a store.

Challenges we ran into

There were two main challenges during this project. The first one is mainly because of the lack of individual’s mobility data during lockdown. Therefore, we had to do lots of research in order to understand what are the key variables that make humans go out; also, approximating how important that variable is for the decision, using the Machine Learning model. Second, the development of the application itself. Having only 9 days to build it, and so many features to add, it was challenging the collection of every API, and to make them work as we needed.

Accomplishments that we are proud of

As a team, we are, first, proud of having implemented a two-front platform that will be able to help two admins to manage their establishments, and users to have more comprehensive data about the place they will head to. Secondly, the development of the application itself is a high accomplishment since none of us was a Flutter expert. And third, a great teamwork. During lockdown one might think that communication will be an issue; however, we were more unique than ever. Each one of us knew exactly what to do, when to do it, and how to do it. Been this stated, the key for a good submission.

What We Learned

How to get things done. Ahmed Ashkar, CEO of Hult Prize said, “Solving the world’s most pressing challenges is not just the right thing to do, it is also a good business”. We learned that getting together our skills can make a difference when developing and deploying a project. If you are working for a purpose, do not try people to follow you; instead, make them love the reason why you do it. With this team we learn not only how to work together, but also how to be passionate about a certain project. Also, we learned that everything we do can be remarkable, if you start by questioning yourself why, then how, and lastly what you are doing.

What's next for Test

Please download the Android APK to test our project.

  • For the user testing, please register (please do not use special characters). Password must be at least 6 digits long.
  • For the Admin testing, use the following loging: User: Password: 123456
  • Check the live demo here
Share this project: