Data Scientist

hackajob on-Demand is currently partnering with an AI startup company to help them hire the best talent. At on-demand, we match and speak with exceptional talent like you and provide insights into the problem they are looking to solve and the interview process.

Role: Data Scientist

Opportunity: Contract

Based: Remote

mozaic is the world’s first global split payment platform for creators.Whether you need to pay collaborators in other countries, recoup expenses before payouts or automatically split distribution royalty earnings with your team – mozaic makes it easy with simple Smart Contracts and flexible APIs for music distributors.

The problem statement from its co-founder and the project

I have music being monetized in multiple places but the challenge is that when I gather all of my sales reports at the end of the month they are very different from each other, so it is hard to see how much money a single song earned across all of my channels.

I would like to easily combine all of the files and see how much I'm making on any given song.

Here are some of the challenges that make that hard:

- Some of the files are in CSV format, some are in Excel and some are in PDFs.

The column names are often different. A song title may call a "song", "title", "asset", "item", "work title", "work" or "project".

- Different sales channels have different encodings: some have only ASCII characters whereas some are UTF-8 so the song names or music artist names across the files may not match directly. Or there may be typos for any number of reasons. For example, "Beyoncé" vs "Beyonce" vs "Beyonec"

- Those same problems can apply to any other columns such as columns relating to the sales channel i.e. "Spotify" vs "Spotify USA" or "Apple Music" vs "Apple"

- Some of these files can be very large, in the millions of rows

An Ideal Solution

The ideal solution would allow me to upload all of my various file formats.

It would link or merge those files into a single format based on fuzzy matching Artist, Song, etc.

It would have to be a near-perfect matching rate since I am using this solution to calculate revenue by song, accuracy is key.