Personalized Action Suggestions in Low-Code Automation Platforms

International Conference on Software Engineering (ICSE) 2023 - Industry Track |

DOI

Automation platforms aim to automate repetitive tasks using workflows, which start with a trigger and then perform a series of actions. However, with many possible actions, the user has to search for the desired action at each step, which hinders the speed of flow development. We propose a personalized transformer model that recommends the next item at each step. This personalization is learned end-to-end from user statistics that are available at inference time. We evaluated our model on workflows from Power Automate users and show that personalization improves top-1 accuracy by 22%. For new users, our model performs similar to a model trained without personalization.