Abstract
Traditional data analytics for television applications rely on manual, dashboard-centric models that track historical user activity. These traditional data analytics for television applications fail to explain underlying user behaviors or to correlate multiple simultaneously launched features with core business objectives. To address these limitations, the disclosed technology describes an autonomous analytical system that utilizes a distributed client-server architecture that hosts a hierarchical multi-agent backend built on an orchestration framework. A supervisor agent coordinates specialized domain engines to execute targeted statistical, performance, and viewership analysis algorithms over read-only television application databases. Communication under strict End-User Credential (EUC) zero-trust security and under an adversarial multi-iteration reviewer agent quality control loop. between a front-end interface and back-end agents is managed asynchronously using remote procedure calls (RPCs) to prevent blocking of a user interface of the television application and client connection timeouts during long-running structured query language (SQL) database queries. The disclosed technology automates data synthesis to transition the television application data platform into an active insight engine. The active insight engine can accelerate decision making velocity from months or weeks to minutes and can reveal complex, contextual correlations.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Mittal, Neetesh; Kushwaha, Vivek; Yadav, Vivek; and Theckyam, Saravanan Santhamoorthy, "Artificial Intelligence Driven Insights and Predictions for a Television Application", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/11473