The world’s largest pizza retailer by sales is monitoring the effectiveness of its enterprise artificial intelligence (AI) models in real-time.
Domino’s Pizza is centralizing its AI and machine learning (ML) efforts on the Datatron platform. By leveraging Datatron, Domino’s seeks to automate and standardize the deployment, monitoring, management, governance and validation of its AI models for workflows in areas such as in-store operations and customer experience.
Domino's was eager to have the ability to monitor models in real-time and understand how predictions were changing over time, so it could refresh AI models with new data. The company also wanted to minimize the involvement needed by its data science resources as new models were being rolled out.
Recognizing this, Domino's relied on Datatron's smart visualization capabilities to create a system to operationalize its models in production and maintain them on a continuous improvement cycle. In this way, the retailer can constantly monitor and update models and create a birds-eye, multi-level view of key metrics to monitor how its models perform in production.
As Domino's continued to scale its internal data science team and business use cases leveraging ML, the company realized the importance of the operations work involved in taking the models to an enterprise-grade production environment. By integrating Datatron's API with multiple business applications, Domino's is able to improve labor scheduling by forecasting the in-store labor required to meet service goals.
In addition, Domino’s can streamline vehicle routing by creating efficient routes for drivers and refine identifying the locations for building new stores. Through automation and standardization of its ML operations Domino's increases the efficiency of managing multiple models at scale in order to optimize outcomes, reduce the efforts required by data scientists and other IT resources, and identify areas of growth.
And by utilizing the Datatron solution and the open-source, Microsoft Azure-based Kubernetes virtual machine orchestration API, Domino's is also able to allocate resources dynamically based on demand, saving both cost and resources.
"Machine learning models can provide significant value to an organization in several business applications, but without a solid ML operations pipeline, that value cannot be truly realized," said Zack Fragoso, manager, data science & AI at Domino's. "We use Datatron as our enterprise machine learning operations and governance platform for mission-critical projects."