SESSION

DeepBrew 2.0 Starbucks ML Platform

OVERVIEW

EXPERIENCEIn Person
TYPEBreakout
TRACKData Science and Machine Learning
INDUSTRYRetail and CPG - Food
TECHNOLOGIESAI/Machine Learning, MLFlow, Orchestration
SKILL LEVELBeginner
DURATION40 min

Introducing our Enterprise Managed Platform, a comprehensive solution designed for the efficient operationalization of AI/ML solutions at scale. This platform employs automation across key stages of the AI/ML lifecycle, offering Model as a Service (MaaS), Data as a Service (DaaS), and Function as a Service (FaaS). With a commitment to establishing engineering standards and best practices, the platform provides managed care for applications, models, data, and feature pipelines. Leveraging Databricks for compute, orchestration, and model management, the platform leverages Databricks MLflow for tracking and model registry. Models are exposed as REST APIs using Azure Kubernetes and Azure APIM for API governance. For monitoring and alerting, the platform seamlessly integrates Datadog and PagerDuty. The platform provides essential configuration management and helper methods for registering artifacts and models and capturing logs for evaluation and retraining.

SESSION SPEAKERS

Vikas Vennavali

/ml engineer lead
Starbucks Coffee Company