HPE Ezmeral ML Ops recognized by four industry analysts

[RichardHatheway]

Great news!

In the last few months, HPE’s industry-leading machine learning (ML) solution, HPE Ezmeral ML Ops, has been recognized by four industry analyst firms. This is great validation of our solution and indicates that customers and analysts alike recognize the value HPE Ezmeral ML Ops provides.

Where was HPE Ezmeral ML Ops recognized? Here’s a quick highlight, which I’ll cover in each section below.  

• Constellation Research – HPE Ezmeral ML Ops listed as a Solution to Know
• Gartner – HPE Ezmeral ML Ops listed as a Representative Vendor for DSML Engineering Platforms
• Gartner – HPE Ezmeral ML Ops included in the Gartner for IT Leaders Tool: Vendor Identification for Data Science and Machine Learning Platforms
• OMDIA – HPE Ezmeral ML Ops was included as a Major Vendor’s Solution for deploying and managing AI models

Constellation Research

Constellation Research is a Silicon Valley technical research and advisory firm providing strategic guidance to companies seeking to transform their businesses through the early adoption of exponential technologies. Constellation publishes the Constellation ShortList^TM that evaluates vendors in a variety of categories that are relevant to early technology adopters. One of those categories is MLOps, which they define as “How to efficiently develop, test, deploy, and maintain machine learning (ML) applications in production.”

In the 2022 Constellation ShortList MLOps report (published 16 Feb 2022), HPE Ezmeral ML Ops is listed as a Solution to Know. This means our product meets Constellation Research’s threshold criteria. To be included in this report, Constellation evaluated more than 20 vendor solutions in this market. Inclusion in this list is based on the following criteria:

• Support for multi-cloud environments. The chosen MLOps tool/platform should support—at a minimum—AWS, GCP, and Azure.
• Its ability to allow and manage the deployment of ML models on Kubernetes with ease.
• Connectivity to as many commonly available data sources as possible.
• Integration with major continuous integration/ continuous delivery (CI/CD) and DevOps tools.
• Support for the following model capabilities:
  • Model versioning
  • Model storage
  • Model training
  • Model deployment
  • Model validation
  • Model monitoring
  • Model registry
  • Feature engineering
  • Model serving
  • Model governance

Constellation Research also uses information from client inquiries, partner conversations, customer references, vendor selection projects, market share and their own internal research to determine what products will be included in the list. HPE Ezmeral ML Ops was included, along with product offerings from AWS, Google, Microsoft, and seven others.

The solutions that are included will be in Constellation’s future market overview deep dive MLOps report. The report is targeted primarily at Chief Analytics Officers, Chief Data Officers, CIOs, CISOs, and Chief Privacy Officers. 

The report may be downloaded for free by visiting the MLOps ShortList page.

Gartner

Last year, HPE Ezmeral ML Ops was listed as a Representative Vendor for ModelOps in the 2021Market Guide for AI Trust, Risk and Security Management.

This year, Gartner has included HPE Ezmeral ML Ops as one of 40 vendors in the 2022 Market Guide for DSML Engineering Platforms. In this Market Guide, HPE Ezmeral ML Ops is listed as a Representative Vendor for DSML Engineering Platforms.

Per Gartner, the market definition of a DSML engineering platforms is that they:

“…consist of a core product and supporting portfolio of integrated products, components, libraries and frameworks (including proprietary, partner-sourced and open-source) for the development and operations of machine learning solutions integrated with typically complex, innovative and highly scalable applications. These solutions are engineered by personas who have deep technical expertise in data science and machine learning or have other skills in digital technology, such as data, software or system engineers. The platforms provide a code-centric user interface, using a variety of programming languages. To boost productivity, they also facilitate composition and automation through visual interfaces and through open APIs.”

Gartner also states:

“DSML engineering platforms have the primary purpose of developing ML models that can drive critical business systems such as credit approval, predictive maintenance, medical diagnosis and fraud detection. In order to do this, DSML engineering platforms have evolved from supporting a core data science audience with code-driven model development to now also supporting data engineering, application development and infrastructure roles. DSML engineering platform development has been driven by the need to enable collaboration between these roles, including their activities and tasks, to effectively deliver ML systems. The focus of these platforms has shifted in all areas of model development. Figure 1 illustrates this shift by summarizing the main characteristics of a DSML engineering platform.”

In the 2022 Market Guide, Gartner also indicates that MLOps platforms is the number one conversation driver for the MLOps social media category in 2021. This indicates how important it is to customers to have an MLOPs platform that provides the functionality they need.

Gartner’s inclusion of HPE Ezmeral ML Ops recognizes that the HPE solution is “…identified under the scope of the emerging data science and machine learning engineering platforms market.”

Gartner

Gartner publishes an annual Gartner for IT Leaders Tool: Vendor Identification for Data Science and Machine Learning (DSML) Platforms. This tool is used by enterprises to evaluate and select providers by comparing significant features.

Gartner defines a multi-persona data science and machine learning platform as a cohesive and composable portfolio of products and capabilities, offering augmented and automated support to a diversity of user types and their collaboration. The primary aim of a DSML platform is to create value through democratization. This is achieved by bringing the power of DSML to a wider nontechnical and technical audience, while hiding complexity “under the hood” by automation and augmentation throughout all phases in the DSML development and operationalization process. Increasingly, this is complemented by offering additional analytics capabilities for business intelligence, visualization, and exploration.

In the most recent version of the tool, HPE Ezmeral ML Ops was included as one of 67 vendors. Gartner identifies the vendors to include in the listing through market research and interactions with their clients.

Vendors’ information is presented in the following categories:

• Collaboration
• Infrastructure
• Openness, performance, and scalability
• Pre-canned solutions / accelerators
• End-to-end data science process
• Data access and preparation
• Data exploration and visualization
• Model development
• Other advanced analytics
• Operationalization
• Governance and responsible AI

OMDIA

Omdia is a global research firm that provides research, consulting, and insights to their clients. In their recent Technology Analysis: AI Edge Platforms report (published 7 April 2022), Omdia evaluated solutions that help customers deploy and manage AI models to the edge. HPE Ezmeral ML Ops was included as a major vendor’s solution that was analyzed, along with hyperscale providers, software vendors, open-source project, and server OEMs.

Omdia’s view of MLOps is as follows:

Two factors, the need for a developer-friendly interface that permits users to get productive quickly and the degree to which nearly all players build using their choice of closed and open-source technologies, are leading the major platforms to resemble each other strongly. Almost all offer an application for exploratory data analysis, experimentation, and model development that also has MLOps features so the user can orchestrate the deployment process from within the same interface. These generally assume and even enforce a model in which AI/ML inference is packaged as an HTTP microservice in a container, relatively independent of the business logic in whatever application it supports.                  

Increasingly, they also offer support for automation, repeatability, and model management. At the same time, almost all platforms use Kubernetes and its AI-specific extension Kubeflow for the underlying infrastructure and offer this directly to users that need more low-level customization. This helps a lot with their own integration challenges.        

Omdia’s observation on HPE Ezmeral ML Ops included the following:

The upstream side of HPE’s offering is a full-featured ML developer environment with extensive MLOps support, based on the same open-source projects we have already encountered, including Kubeflow for training and evaluation pipelines, Apache Spark for distributed data processing, and AirFlow/MLFlow for ML training. The system offers HPE’s own notebook environment. It also supports imported open-source tools or downloads from the Ezmeral Marketplace app store and integrates with common DevOps tools for version control, testing, CI/CD, and the like.

The key selling point is probably that data management and higher-level applications development are supported as first-class citizens within the same platform. HPE likely sees this as a way of selling hardware through the software/services channel rather than giving away software to smooth adoption of its hardware.

HPE Ezmeral ML Ops – the product

HPE Ezmeral ML Ops addresses the challenges of operationalizing ML models at enterprise scale by providing DevOps-like speed and agility, an open-source platform that delivers a cloud-like experience, and pre-packaged tools to operationalize the machine learning lifecycle from pilot to production. It supports every stage of the ML lifecycle—from data preparation to model build, model training, model deployment, collaboration, and monitoring. HPE Ezmeral ML Ops also provides enterprises with an end-to-end data science solution with the flexibility to run on-premises, in multiple public clouds, or in a hybrid model and respond to dynamic business requirements in a variety of use cases.

To learn more about HPE Ezmeral ML Ops, visit hpe.com/mlops.

Hewlett Packard Enterprise

HPE Ezmeral on LinkedIn | @HPE_Ezmeral on Twitter

@HPE_DevCom on Twitter