Machine Learning can be used strategically to solve business problems that involve analysis of large data sets. Formulating an effective Machine Learning Strategy means establishing a well-articulated business goal that can be solved through data analysis, defining the scope of that goal, identifying the necessary data sets and determining key indicators within the data, and deciding on methods of data ingestion and analysis.
We took a content aggregation platform and trained an algorithm to improve content curation by distinguishing between original content and covers/remakes, and then prioritizing original content in the display order. This process was originally handled with assistance from human curators, but then improved to be handled through machine learning only, in order to facilitate scaling the platform. We also developed a strategy to automatically generate the UI based on the type of content being viewed.
This project involved navigating the legalities of aggregating copyrighted content, developing a business model that incentivized the use of the platform by affiliates, and go-to-market strategies that required managing complex B2B partnerships.