A mechanism is proposed for predicting a subset of users for a feature rollout in a cloud storage system. The method includes identifying a first set of users having user accounts associated with the cloud storage system on a random basis. The method also includes providing access to a new feature of the cloud storage system to user accounts associated with the first set of users. Further, the method includes determining, for each user in the first set of users, user interactions with the new feature. The method furthermore includes predicting a second set of users who are likely to be satisfied with an improved version of the new feature. The second set of users is larger than the first set of users. The improved version of the new feature reflects the user interactions by the first set of users with the new feature. The method includes providing access to the improved version of the new feature to user accounts associated with the second set of users.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Procopio, Michael, "INTELLIGENT DATA-DRIVEN USER POPULATION SUBSET SELECTION FOR PROGRESSIVE FEATURE ROLLOUTS", Technical Disclosure Commons, (September 24, 2018)