Disasters are chaotic. Non-profits scramble to deploy resources and volunteers. 160 million people die or are fatally injured because resources were not reached to them in time. At the same time, $60 million worth of non-profit resources are wasted every year. What gives?
Reach the victim with the right resources at the right time. Collaborate if necessary. How? Information management. This is why we built Relief.net (Relief Network)
How we built it
We thought about the problem and thought about it some more. We debated, argued and fought over the litany of ideas. We worked closely with our non-profit to build a beautiful, accessible user interface and used machine learning in thoughtful ways. The victims can use the app to request resources, and can use the voice-activated chatbot to do the same. The tickets generated by victims are assigned to the right non-profits, who can collaborate and share resources to solve the ticket. We have analytics built in, and an option to involve community citizens in relief efforts though public tickets and donations for projects.
Challenges we ran into
Thinking from the victim perspective in chaotic situations, and building features that have the most impact for them.
Accomplishments that we're proud of
- A prioritization model for tickets that improves w/ the help of machine learning
- A gorgeous, user-friendly and accessible user interface
What we learned
We can use machine learning and thoughtful interactions to improve non-profit collaborations and impact.
What's next for NPO System
Scaling it! Working with real world messy data :)