Bringing foresight to API management to leverage digital marketplaces

Differentiation and integration

As digital marketplaces mature, they evolve, and we are now well and truly into the era of digital marketplace 2.0, where, as we explored here, differentiation is the name of the game. In a highly competitive world, a digital or physical item or service is much the same, regardless of where the customer buys it. The only differentiators here are price – and a race to the bottom isn’t a sustainable business model – and customer experiences, which now are generally similar too.

The way to differentiate is by offering something more to a particular community that they cannot get elsewhere and finding ways of making better use of existing assets. Data-as-a-Service (DaaS) is a perfect fit for both, but it demands a new way of looking at things. Before the astounding rise of the digital natives, data was typically mostly seen as a byproduct, just as IT was viewed as an unavoidable overhead.

The digital natives turned this on its head. Their IT estates are their businesses, and everything they do is driven by data that provides hard evidence of how customers behave, their needs, and their preferences. It also allows these companies to make predictions to better serve their customers and experiment until they get new offers right. And they also sell selected, anonymized data to third parties, such as advertisers in the case of Google and Facebook.

There is immense scope to adopt and adapt these models in industry verticals, where insightful data and highly trained specialists are hard to come by. Schneider Electric, a leading provider of energy management devices and solutions, is a good example of a DaaS pioneer. More than a decade ago, it realized that IoT and the industrial internet would profoundly impact traditional industrial sectors, such as its own, and started thinking about how it could best leverage the new opportunities.

Schneider chose Torry Harris Integration Solutions (THIS) to help it transform its IT applications landscape to benefit from these huge shifts. A big factor in THIS’ favor at the RFP stage in 2014 was its track record in very large-scale, highly complex transformations that involved bringing massive amounts of legacy into the fold.

APIs support diversity

Schneider is a multi-billion-euro corporation headquartered in France and, at the point at which THIS got involved, had acquired 600 companies across 150 countries, along with their IT estates. Rationalizing such a huge, diverse, complex estate is extremely difficult and complicated further by the 12 or so acquisitions the company expects to close annually. Hence Schneider’s approach was to maintain the diversity of the estate, only replacing tech or applications when they became obsolete or were not functional.

Now the big question remained, how to extract value for the gold mine of data it was sitting on, buried deep in many siloes right across the organization? It is important to note that from the start, Schneider’s primary goal was to find new revenue streams by delivering more value to customers rather than operational efficiency.

It wanted to enable cross-domain collaboration and to share the assets of key platforms, such as within the Information, Process, and Leadership (IPO) unit and strategic divisions, including Digital Customer Experience and Global Solutions. At the start of the transformation, all the integration programs within Schneider were point to point, driven entirely by project requirements. This meant they could not be reused, time to market was slow because the integrations had to be redeveloped, and maintenance was complex because data was held in middleware to determine business logic.

Putting the backbone into integration

In just over a year, THIS worked with Schneider’s Enterprise Architecture (EA) team to create an integration backbone for its global IT systems to open up Schneider’s application programming interfaces (APIs). It decided on an API-driven solution because:

• They are more open and well documented
• Can often be self-provisioned, which makes them better suited to consumption by partners, and,
• Because of their strong business affinity, the APIs themselves can be marketed as products.

On the business side, Schneider’s EA team and THIS worked with the firm’s business strategy leaders to identify monetization opportunities for each unit.

On the technical side, to map its progress, Schneider benchmarked its goals against a telecom provider (telcos are heavily reliant on APIs) to understand the future state of its business better. Like that provider, it wanted to adopt concepts such as usage-based billing and paid-for software upgrades for its energy management products. It also planned to connect devices to gather data, analyze customers’ usage patterns and visualize data so customers can literally see ‘what good looks like’ and what they need to do to keep ‘looking good’ in different scenarios

Customers would be able to access these services via an API app store to drive incremental revenues.

Making it happen

The EA team and THIS aligned domains and sub-domains within the IT organization and set up architectural governance and management principles for the application portfolio. A critical success factor for Schneider is that APIs are reused at every opportunity and deployed and maintained consistently. Piecemeal upgrades and customizations damage interoperability and disrupt the operational and business model.

The API management platform comprises:

An API store where customers can see which APIs are exposed by various businesses and units across Schneider, plus detailed documentation about each one and how to consume them

+

A dedicated sandbox allows customers to try them out in a test environment

+

An API management dashboard provides statistics to support and inform the developer community, including visibility of who is using which APIs and for what purposes, in real-time

Initially, the API platform catered mainly to internal consumers and helped improve cross-domain consumption of services and APIs and the reuse of IT assets within the enterprise, lowering cost and time to market. The platform also provides a method of monetizing the APIs through internal metering and chargeback. For example, the integration framework allows Schneider's Global Solutions (which provides digital software across the business) to build solutions like subscription-based billing, a customer information database, and a product information database, being leveraged to generate new revenues from digital products.

THIS helped develop the working API model and the publication of the full details on its operating elements, principles, best practices, and contextualized patterns. This empowers the EA team to progressively expand, refine, and replicate this "reference architecture" across the enterprise.

Next steps

In phase 2, THIS provisioned the services and API stack on which the business-built applications, and improved visibility of APIs through an API discovery experience. This collates and displays the Schneider Service Catalog data in the API app store, categorizing services (by tagging) and the associated usage of APIs to give the providers better visibility and control of their API. Phase 2 also ensured an improved quality of service and monitoring of the health of APIs against service level agreements (SLAs) and any breaches in compliance with them.

In summary, the large-scale integration enabled Schneider to:

Reporting on the project, Ovum ( ) concluded that, “Incredible savings came from the internal and external reuse of services from an existing catalogue. The potential savings are slated to be in the range of millions of euros by 2020. The time to market for a project due to reusability, as well as the already established governance and processes, empowers accelerated digitalization in Schneider IPO.”

Productizing expertise

THIS’ has extensive, in-depth experience and expertise gained from working with partners in large-scale integration projects. In the era of digital marketplace 2.0, THIS has productized that expertise and experience in the shape of 4Sight, which is an API-driven tool, and built to work with any source of data: logs, databases, streaming data, and more – to deliver intelligent insights to users.

The framework generates widgets that can be embedded into any existing application, thereby seamlessly bringing analytics and prediction capabilities to an application without massive disruption and retrofitting. The configurable machine learning models allow users to deploy 4Sight for multiple purposes, including AIOps and event predictions.

In April 2022, THIS’ 4Sight won a Computing DevOps Excellence Award in the Best AIOps/MLOps Tool category. This was for its outstanding analytics and predictive capabilities developed to address the problems of fragmented data.