HPE Haven OnDemand and Microsoft Azure Machine Learning: Power Tools for Developers and Data Scientists

While both HPE and Microsoft machine learning platforms offer numerous possibilities for developers and data scientists, HPE Haven OnDemand is a diverse collection of APIs for interacting with data designed with flexibility in mind, allowing developers to quickly perform data tasks in the cloud.

Prediction API

Under the ‘Try It’ tab of the Prediction API, we are now able to make predictions with our model. Again, by simply providing an input source, the class attribute to be predicted, and the name of the pretrained prediction model (from the previous step), HPE Haven OnDemand is able to predict class attributes and return the results as a JSON document, flexible enough to be further processed in any way a developer or analyst sees fit.

HPE Developer Site

For more information, and a detailed tutorial on the above prediction process using HPE Haven OnDemand, see here. For more on developing applications, see the HPE Haven OnDemand Developer site.

While this article has focused on the developer portal, it should be mentioned that havenondemand.com currently hosts the following 2 solutions-as-a-service, start-to-finish browser-based solutions built on the underlying HPE Haven OnDemand APIs, designed for business users:

Of note, HPE Haven Search OnDemand is also fully-integrated with HPE Find, an open source search tool, allowing developers and organizations the flexibility of using the HPE OnDemand APIs to further customize the search experience using the HPE Haven OnDemand APIs.

Microsoft Azure

To provide some context for HPE Haven OnDemand and its place in the analytics ecosystem, we have a quick look at performing a similar task with Microsoft Azure Machine Learning. Microsoft Azure is a cloud computing platform for managing various Microsoft and third-party applications and services, from data processing to storage to analytics and beyond. A leader in the IaaS and PaaS space, Azure is very much a full-fledged collection of infrastructure and platform services, by design; however, we are specifically interested in comparing HPE Haven OnDemand with Azure Machine Learning.

Performing the same prediction process in Azure Machine Learning means building a complete machine learning pipeline. It is a more time-consuming process, but also offers a much richer set of tools and services. Azure Machine Learning provides a web analytics app studio for data scientists and developers, compared to HPE Haven OnDemand’s simple, yet straightforward, APIs and online console.

A glaring difference between platforms is that Azure Machine Learning supports Python and R, whereas, as previously noted, HPE Haven OnDemand allows its APIs to be called using nearly any programming language. As a Microsoft product, Azure Machine Learning does come with the usual high-quality technical documentation and resources; however, considerably more time must be spent acclimating to the product before becoming productive, by way of videos and tutorials, before making your first prediction. Contrast this with HPE Haven OnDemand, where a one page tutorial lets you hit the ground running. HPE Haven OnDemand represents simplicity; navigation, getting help, and getting started are all easily accomplished via its website. Microsoft’s conflated Azure branding and products may provide a rich experience for those who have learned the ecosystem, but can be time-consuming and intimidating for those who have not.

While outcomes between the services for this particular task are comparable, the time required to get to the same point is quite different. At a minimum, a pair of simple cURL calls or browser-based interactions could return the same results as an extended period of time spent in Azure Machine Learning.

The reason again is simple: HPE Haven OnDemand is designed as a quick and flexible set of APIs to interact with data anytime, anywhere, while Microsoft’s Azure Machine Learning platform is a workbench for managing end-to-end pipelines. Use cases differ, as do the tools most appropriate for said use cases.

One last point of comparison is pricing. Determining Azure Machine Learning pricing is a much more complex process than is HPE Haven OnDemand pricing. Whereas Azure pricing includes a web of explanatory pages, an online calculator, and numerous Azure family configuration options, HPE Haven OnDemand has a single page dedicated to pricing, which clearly enumerates a small number of packages and their associated monthly costs.

While not the focus of this article, it is worth mentioning that Microsoft’s Project Oxford may provide a more fitting, if less well-known, comparison to HPE Haven OnDemand. A small collection of APIs for vision, speech, and language tasks, Project Oxford, unfortunately, lacks a prediction API, and as such was unsuitable for this particular comparison.


Both HPE and Microsoft machine learning platforms offer a wealth of possibilities for their end users. Developers and data scientists now have a rich set of enterprise-grade machine learning tools at their disposal to tackle a wide array of use cases.

HPE Haven OnDemand is a diverse collection of APIs for interacting with data. Compared to Microsoft Azure Machine Learning, a full-featured machine learning app studio for managed pipelines, HPE Haven OnDemand is designed for flexibility, allowing developers to quickly perform data tasks in the cloud via a web interface, HTTPS, or cURL calls, accessible from nearly any programming language. HPE Haven OnDemand truly lives up to its promise of interacting with data “anytime, anywhere.”


Note: the article was commissioned by HPE, but written by an independent KDnuggets expert.