In a highly competitive industry such as Telecommunications, adopting a data-driven company-wide approach is no longer an option. The creation of an Artificial Intelligence Center of Excellence and the transition to a Cloud development environment allows to better manage corporate data and enjoy the benefits of AI models to support decision-making, across all business units.
The proposed approach for the AI CoE setup is structured in two parallel streams. The first one, the Data Strategy, aimed at the identification of possible use cases over all business units and the definition of a roadmap of AI initiatives prioritized with respect to business strategy and economic benefits.
The second stream consisted of the definition and setup of a cloud environment, leveraging on AWS components. AWS LakeFormation has been used to create the data platform for collecting and managing data consumed by the AI models.
As Analytics Platform, SageMaker is the enabler of all the activities of data analysis: ranging from data exploration to reporting, passing through data manipulation, model training and parameter tuning. Furthermore, SageMaker Studio brings all the steps of Machine Learning development lifecycle in one place, providing a single web interface for creating notebooks and instances on demand.
As result of the Data Strategy, the multi-year cross-department AI CoE roadmap has been defined and 7 priority initiatives were selected for the first year. Moreover, the AWS cloud environment has been set up. The designed architecture grants flexibility and scalability of the development and execution of AI models.
100+ uses cases evaluated across 12+ Fastweb departments
Cross-company multi-year roadmap defined, with 7 use cases in the first year
Centralized and efficient management of AI development, artefacts, and data