Abstract
Agentic workflows can automate many user journeys. However, when a user journey involves choosing between different options, e.g., dietary preferences, the AI agent may be inadequate in automating the user journey, requiring manual actions and/or user inputs to make a choice. This disclosure describes techniques that automate agentic workflows by permitting an AI agent to access user preferences from a user profile prior to performing a corresponding task. With user permission, user preferences are obtained via explicit input, user browsing history, transactions, etc. and are stored in a user profile that is accessible to AI agents. When summoned to perform a task, the AI agent categorizes the task into a particular type (e.g., food-related). The AI agent is configured to access the user profile in accordance with user-selected permissions to obtain the user preferences (e.g., dietary preferences) relevant to the task. Thus, the agentic workflow is implemented with knowledge of relevant user preferences and can provide more complete automation of tasks that involve parameters related to user preferences.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Azose, Benjamin and Gupta, Akancha, "Retrieving Personalized Content Using AI Agents and User Profile Information", Technical Disclosure Commons, (February 10, 2026)
https://www.tdcommons.org/dpubs_series/9311