We were inspired by the success of numerous up and coming video sharing applications that have been entering the market for years. From snapchat to vine to beme; they all revolve around sharing who you are with the people around you.
But they are all missing something. Every _ popular _ user on vine and snapchat super-star are incredibly similar. They become popular making short crowd mentality inducing videos that really carry no substance. This is where prompt steps in.
What it does
Rather than a user standing in front of a camera and talking to their audience with pre-prepared speeches, prompt will _ prompt _ you with a question that you will have a few seconds to think about before which you will have to answer it. This might seem a little bit intrusive at first, but it levels the playing field in the global community, and prevents users from putting up fake walls around themselves just for their audience on their social app.
With this comes added benefits. You can use this application as a private video journal and make as many personal entries as you like. This lets people "scroll back" on their life and reflect upon what they thought, how they matured, and enjoy nostalgia that the app is sure to bring.
How I built it
The front end was built in swift. The UI is not overly complicated and it is clean and easy to use. We took advantage of the numerous tools that xcode 7 has to offer which made the application design relatively painless. The back end however was the real meat of the project. We decided to build the backend in a way that we could essentially scale to millions of users instantly. We built the backend on top of kubernetes and amazon ec2 which allows us to dynamically stir up more servers when under high load. This is theoretically infinitely expandable. IN addition AWS and kubernetes, we took advantage of many cloud based tools which we deployed to our our cluster running on AWS.
Challenges I ran into
This project was built to scale larger than anything we had previously worked on. We used best practices from industry giants like Netflix and Apple to build our system. This was not easy to say the least. After reconfiguring the cluster over 100 times in 24 hours due to various routing and dependency issues. Finally we were able to call the cluster stable.
Accomplishments that I'm proud of
The sheer magnitude of the project and how quickly we completed it is beyond comprehension. We took on this project as an exercise in building a real world scalable service. And we completed that goal despite none of us ever using swift or ios development before. And 2 of us never having attended a Hackathon before.
What I learned
This whole experience taught us a lot. For one, ios development is very interesting and we can definitely see ourselves building ios applications again. But the real lesson was in the sheer magnitude of work and understanding it takes to build a scalable system like we have.
What's next for Prompt
We plan on implementing many more features from facebook integration, to prompt suggestions. Our list of potential features is ever growing and we would love to take this application to a point where it is just as much a household name as snapchat or instagram is today.