Your browser is not supported. Please update it.

22 November 2019

Stealthy Cloud Information Model threatens Open Data Initiative


Any hopes of an uncontested cloud-based data standard now appear shaky, thanks to the announcement of the Cloud Information Model (CIM) project. The CIM looks set to compete with the Open Data Initiative (ODI), which launched just over a year ago in September 2018. Both want to solve the problem of cloud-level data interoperability, but we’ve seen how IoT-type standards battles play out.

The irony here, as is always the case with standards that seek to unify; as both the ODI and the CIM recruit followers, the desired objective of pooling and standardizing as much data as possible becomes ever more distant. It is not clear why the two organizations could not simply merge now and pool their resources, nor is it clear why the CIM membership decided it had to form an entirely new group instead of working within the ODI.

The CIM’s planned output is a data sharing standard led by the Linux Foundation’s Joint Development Fund. It was announced last week, although the fanfare was so muted that the three major partners – Amazon, Genesys and Salesforce – almost seem embarrassed by its creation. The only press release was on the Linux Foundation’s website.

Meanwhile, the ODI announced this week that it had recruited Software AG to its Partners Advisory Council.  This is a good gain for the more established standard which was established by Microsoft, Adobe and SAP.

We expect that the arrival of CIM means that data sharing standards will eventually become a war of attrition. Both will drive to recruit the growing number of companies that are waking up to the value of freely transferrable, convertible and accessible data from a variety of internal and third-party sources, all processed in the cloud.

Until there is a victor – and who knows when, or even if, that will occur – both standards are likely to impact the way in which data from the IoT is processed, pooled and utilized. Already there are companies that have ties to both the CIM and the ODI which means we are still some time away from entrenched tribalism. Could the two models end up converging?

Both the CIM and the ODI promise businesses the ability to freely use all accessible data by releasing it from various silos and standardizing it – an offering highly valuable to IoT implementations. Effectively for any type of job you want doing – say sales order processing – there is a subset of the data model that deals with that subsegment, and no-one has to re-invent it and multiple app providers can use the same one.

The CIM is newer, formed by AWS, Genesys, and Salesforce in partnership with the Linux Foundation’s Joint Development Fund. The website summarizes the standard rather neatly – ‘an application-agnostic data model that simplifies integration’.

Much like the ODI, the CIM promises to break down the ‘barriers to innovation’ that are formed by siloed data sets and incompatible data management platforms between services.

The CIM currently provides a standard for data relating to six ‘subject areas’, or major business concepts. These are Payment, Payment Method, Product, Sales Order, Shipment and Party – the trade actors involved and the relevant means of communication.

The standard is enabled by MuleSoft’s open source modeling technology, which comprises of Anything Modeling Language (AML) – a declarative language for defining metadata documents – and AML Modeling Framework (AMF) – an open-source library that can analyze and validate any metadata document.

In such early days, the CIM has yet to attract followers beyond its founding partners, but considering one of them is AWS, this is hardly a poor start.  Use cases already in action are AWS’ Lake Formation service, which uses the standard to create secure data lakes with data from a multitude of sources, as well as ‘data warehouse’ Amazon Redshift.

The CIM’s two other partners have also put the standard to use. Salesforce’s CRM platform Customer 360 operates according to CIM – a kick in the teeth for Microsoft, as it was expected to power Customer 360 with the Azure platform. The CIM standard is also apparently used for Genesys Cloud and Genesys AI, two platforms that allow companies oversight of customers and employees by combining engagement data from all communication within an organization.

Aside from the CIM, AWS has recently been engaged in other data sharing initiatives. Last week the company announced the AWS Data Exchange, a platform in which customers can find and utilize third-party data available on the AWS Marketplace.

The Exchange holds 1000 products containing data from over 80 ‘qualified data providers’ such as Reuters, Change Healthcare, and Foursquare. The available ‘data lakes’ advertised include 14 billion healthcare transactions from Change Healthcare and 330 million global business records from Dunn & Bradshaw – although how ‘open’ this sensitive data will be is questionable.

By contrast, the ODI is more established. The common data model was announced at Microsoft’s Ignite conference in September 2018. The first major recruitment drive came at the Adobe summit in March this year, where the ODI leading partners – Microsoft, Adobe, SAP –announced 12 new partners which would form the Partner Advisory Council. These initial members of the Council included companies such as Accenture, Cognizant and Sprinklr.

Much like the CIM, the ODI aims to make data sharing between technology vendors’ services more streamlined and convenient by providing a common data format. The ODI’s marketing conjures images of emancipated data silos, flowing into vast ‘data lakes’ from which advanced analytics and comprehensive business insights can operate.

What’s become apparent is that some of the firms on the ODI’s Partner Advisory Council are now partnered with the CIM and/or AWS Data Exchange. Change Healthcare is one of the flagship first users of the AWS Data Exchange, whereas Genesys is one of the three founding partners of CIM.

In a world as murky as the cloud, it is hard to see exactly where battle lines are being drawn. It is highly unlikely that any partner of one standard will cut all ties to a firm which practices the other. For instance, last month SAP announced further collaboration with Microsoft, on an initiative to help businesses move their data to the cloud. SAP co-CEO, Jennifer Morgan, said that “as always, choice will prevail as we recognize that many of our customers also run SAP on AWS and GCP, for example.