Inspiration

The inspiration for the project came from the

What it does

The Reverb Audio Comparison Microservice is a tool that can be used to analyse and differentiate audio signals in an attempt to compare and understand their similarities. It can be used in many applications ranging from music recommendation to audio recognition.

There are multiple aspects to the hack that need to be considered. The Reverb Microservice can be used to offer a solution to enterprise customers to incorporate our core backend functionality with their existing services. To demonstrate this, we have develop a Live Band finding recommendation (using a users spotify account) service to showcase the potential of the platform.

The platform can be monetized through our own api with payments generated by affiliate leads.

How we built it

The neural networks was built using Google TensorFlow. The Demo Gig recommendation service was built using Node.js, Angular.js and Firebase.

Challenges we ran into

There was a lot of time spent on the neural network without access to powerful computers.

Accomplishments that we're proud of

  • Building the model
  • Polishing the demonstration

What we learned

  • We need moar power.
  • To think about monetization as well as functionality.
  • Identifying different ways we can provide a service to our customers.

What's next for Reverb: Audio Comparison Microservice

Improve the model and see how far we can take the idea.

What we used

  • Spotify API
  • Million song data set
Share this project:
×

Updates