Hewlett Packard Enterprise Co. today is expanding its reach into artificial intelligence development with a software platform that supports the full lifecycle of machine learning model construction and deployment using the self-contained software environments called containers.

HPE ML Ops provides for the rapid rollout of machine learning workloads across on-premises, public cloud and hybrid cloud environments. The idea is to enable development teams to employ processes similar to those used in DevOps, the rapid application-building technique that that involves frequent code releases and constant refinement. The result is reductions in model deployment times from months to days, HPE said.

The company is attacking a common problem with machine learning projects, which is a lack of resources and operational processes to deploy them. Gartner Inc. estimates that half of machine learning projects fail to be fully deployed due to lack of operational support. Part of the problem is that machine learning models can be complex and require large amounts of data and computer horsepower, making provisioning both expensive and time-consuming.

HPE figures containers can help. It’s building on top of the EPIC software platform it acquired with the purchase of BlueData Software Inc. late last year. EPIC includes pre-configured versions of popular AI and analytics applications that are packaged into software containers that can be stored in a library and quickly put into service.

ML Ops is said to cover the full lifecycle of machine learning development, including model building, training, deployment, monitoring and team collaboration. Developers have access to self-service “sandbox” environments for tools and data science notebooks. Completed models can be reproduced and reused. Development teams get full visibility across the model lifecycle and security includes multi-tenancy and integration with enterprise authentication mechanisms.

HPE said ML Ops works with a wide range of open source machine learning and deep learning frameworks including Keras, MXNet, PyTorch, and TensorFlow as well as commercial machine learning applications from its ecosystem partners including Dataiku Inc. and H2O.ai Inc.

Pricing wasn’t specified.

Image: A Health Blog/Flickr