IBM Launches Experimentation-as-a-Service Offering

IBM Research has launched a new Experimentation-as-a-Service offering that enables users to create demos of products and services and host them on the IBM Cloud.

Experimentation as a Service is not a new thing; other cloud computing infrastructure providers such as Microsoft and Amazon have taken a crack at this issue. Experimentation as a Service enables users to use experiments to accelerate innovation by providing a cloud environment for demonstrating and verifying new products and technologies.

Indeed, the new IBM Research Experimentation as a Service (EaaS) offering enables “demo creators to define and implement a set of services and UIs that provide an abstraction of the underlying cloud resources (including VMs and Networks), making it easier for individuals to quickly develop, classify and recall complex demonstration topologies and experiments on an as needed basis,” according to a description of the service on IBM’s Cloud Marketplace site.

On July 1, Michael Rowe, a business development executive at IBM Research in Research Triangle Park, North Carolina, tweeted, “After months of hard work. Announcing—IBM Research Experimentation as a Service by IBM,” including a link to the service on IBM’s Cloud Marketplace. That tweet, however, disappeared after eWEEK inquired about the service.

In response, an IBM spokesperson said, “There isn’t really a story to tell around this right now.”

Yet, IBM’s EaaS online user agreement says different. The agreement reads: “IBM operates, through its Cloud Marketplace, an IBM Research Experimentation-as-a-Service offering that is intended for experimentation on and technical demonstrations of new and emerging technologies, which may be updated or modified by IBM from time to time …”

According to the EaaS entry on IBM Cloud Marketplace, the service promises to help users accelerate delivery of demonstrations, save time and eliminate one-off installations with unpredictable results.

EaaS also offers a form-based approach, fine-grained access control and runtime support to help reduce deployment complexity.

IBM also provides a wizard that runs during the construction phase of a demonstration that provides a form-based approach for creating a model of the experiment identifying the various components of the experiment.

According to IBM, before EaaS a company’s sales force was limited to marketing material and presentations to share their product ideas with their customers, due to the high cost of reproducing live, interactive product demonstrations. EaaS changes that, enabling users to provide a more immersive approach in which customers can experience how the product works without installing new software, IBM said.

The EaaS solution comes in standard, large and extra-large editions, starting at $2,500 a month. According to the IBM Cloud Marketplace entry for EaaS, EaaS-Standard costs $2,500 a month and is targeted at technical sales teams. This edition includes support for unlimited users and five servers. It designed for those getting started—testing demos with a small VM footprint.

EaaS-Large costs $4,500 a month and is targeted at technical sales teams that want to create demos with moderate VM footprints. This edition supports unlimited users and 10 servers.

EaaS-Extra Large costs $8,100 a month and is targeted at technical sales teams that want to create demos with a large footprint. This edition supports unlimited users and 20 servers.

IBM later responded to eWEEK with a statement saying that the scientific and IT communities see an ongoing demand for simplifying research collaboration and storing and sharing scientific experiments and demonstrations outside of the lab, in a variety of different locations and on different machines.

“To help meet this demand, IBM is now offering IBM Research Experimentation as a Service to the public through a ‘freemium’ model on the IBM Cloud Marketplace,” said Dinesh Verma, IBM Fellow and Department Group Manager of IBM Research Cloud-based Networks.

Verma said this service was originally developed to support research conducted within two different academic and research consortiums: the International Technology Alliance in Networks and Information Sciences and the Network Science Collaboration Technology Alliance. As the service is considered to be in “beta,” IBM Research will be taking user feedback and continually making improvements and updates to the service, he said.

Meanwhile, as far back as 2008, Microsoft introduced a platform known as the Microsoft Experimentation Platform (ExP), which the company at the time described as a platform that enabled product groups at Microsoft as well as outside developers to innovate using controlled experiments with live users.

“The platform enables testing new ideas quickly using the best-known scientific method for establishing causality between a feature and its effects: randomized experimental design,” the description said.