Abstract
Time tracking apps and manual time tracking can help users track their activity throughout a day and identify practices that can help improve the way they use their time. However, in many cases, such activity tracking may not be accurate and is not performed across multiple devices that a user may use. This disclosure describes techniques, implemented with user permission, that enable users to accurately track their time across multiple devices, obtain insights into their activities, and recommendations regarding device settings, activity parameters, etc. that can help improve their productivity. With user permission, logs and other low-level granular data from individual devices are processed on-device to generate aggregated data indicative of user activity on the device. Data from multiple devices is combined and analyzed using a multimodal artificial intelligence model that generates information regarding activity categories and activity breakdowns, as well as personalized insights for the user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Guajardo, Jaime and Sutasirisap, Daowz, "Context-aware Time Tracking, Activity Categorization, and Recommendation Generation Using Multimodal Artificial Intelligence", Technical Disclosure Commons, (October 16, 2025)
https://www.tdcommons.org/dpubs_series/8734