All charity works are categorised
CherryTea allows users to quickly find events based on their preferences, interests and location
It's easy to participate in something worthwhile
SMS notifications about any changes in participated events
Every day volunteers help people in need all around the world. To be a volunteer you can do as little or as much as you want. However, we shouldn’t forget about the people that surround us. Think not only global issues, but the problems in our communities.
What it does
CherryTea is a mobile application that helps people find
charity work inside their
communities. Our unique recommendation system, created using Microsoft Machine Learning algorithms, allows users to quickly find events based on their preferences, interests and location. Our application provides opportunities for smaller organizations, presents an unbiased list of activities and events, which
doesn’t depend on the size of organization.
How we built it
We have formed a team of enthusiasts from different backgrounds. We have prioritized tasks and set a goal to build an MVP during the
2018 Hex Hakathon. Our mobile app is built with
Kotlin. Behind the app, we have a lightweight server solution with a
Azure ML Recommendation System. Our backend is written in
ASP.Net Core and deployed with
CI/CD to our dedicated server.
Challenges we ran into
We were able to scope the problem well on service level (app, recomendation system, backend). The unfortunate part was absence of appropriate data sets. The more data you have, the better recommendations you get. We didn't find desired information, so we generated fake data to emulate user activity.
Accomplishments that we're proud of
- Our mobile app. Beautiful, solid UX and great visuals. Definitely the eyecatcher of our project.
- Our focus and endurance. Our team literally worked from dusk till dawn without losing its spirit for a moment.
- Machine learning. Being able to build up a model that recommend charity works with thin background on machine learning truly gave us a confidence boost. Even though we had little experience in machine learning we were able to scope the problem wisely and got good results. There were many setbacks, but we always found a solution or some alternative way to overcome those.
- Innovativeness. This is embedded in our team. There were many sudden challenges that we couldn't estimate. We still found a way to go forward.
- Proper DevOps is a thing that you don't want to neglect, even if you are in a high-speed hackathon. When you have solid foundation it feels safer to develop on it.f
- Integration with
Twilio. Volunteers will be inform about any changes in participated activities via SMS.
- Creating an application that is really helpful.
What we learned
We all wanted to learn something new and we truly had that experience. Diving deep into recommendation systems was scary and challenging. Developing modern mobile app after years of desktop development wasn't a walk in a park. With the right attitude and hard work we outdid the learning curves of
Twilio and many other tools that were not that familiar to us.
What's next for CherryTea
In the future, our application will allow any user that needs help to register and get it from compassionate philanthropic volunteers. We are going to improve our system by matching volunteers with people in need faster and more accurately. We believe that CherryTea will make the world a better place.