Movie Night
Movie Night is a feature that focuses on matching diverse movie tastes to ensure users find their ideal movie to watch together. This feature goes beyond conventional recommendations by using AI-driven search capabilities.
Problem
Netflix provides a variety of recommendations tailored to individual users. However, identifying a movie that resonates with the tastes of multiple viewers presents a significant challenge. The vast array of options can transform the anticipation of a movie night into a lengthy decision-making ordeal.
There's a need for a more efficient solution—one that reduces search time and elevates user convenience, ensuring a seamless movie-night experience.
Feature Objective
Match individual users' tastes to ensure a comprehensive movie experience.
Reduce the time taken to find a suitable movie.
Recommend movies based on AI-driven results tailored to user input.
Feature Description
How does it work?
Users can link their profiles and therefore combine their individual tastes in movies.
By comparing their top movies, the system traverses through their recommendations to find the optimal movie suggestions that work for both users.
Users can utilise an advanced AI-search, inputting their specific requirements:
e.g. "A romantic comedy set in Vienna with Ethan Hawke"
The users only get one recommendation at a time to urge them to finalise their decision.
“As a user, I want to be able to link another profile to combine our preferences while also being able to specify our current mood and get accurate recommendations.”
Impact and Benefits
Increase the user experience by tackling the one main issue on a movie night.
Separate the platform from competitors, making it more likely to be chosen as the platform for social viewing or movie nights.
Potentially widen the individual user’s taste and therefore increase their platform engagement!
Implementation
An early prototype was designed, developed, and deployed to explore the potential of this feature. Due to the unavailability of the Netflix API, I utilised mock movie data to ensure the product felt complete. For the AI-powered search mechanism, I used OpenAI's API. It's worth noting that the AI search currently takes a bit longer, so I appreciate your patience during testing.