Optimizing an AI pipeline with Kubeflow
Algolia - 55 Rue d'Amsterdam, 75008 Paris - Wednesday, March 18th, 2020
In this talk, we want to share our findings and experiences on building platform services for managing AI and ML-pipelines in an efficient manner. Using container-based technologies like Docker and Kubernetes we can improve the data scientists’ workflows without the need to know what happens in the background. It’s all about automation and self-service, whether it’s in the cloud or built as on-premises infrastructure. Managing data is also a very important yet often underestimated challenge to cover and talk about as this is the asset from which we generate the most value.