APIs, designed to provide abstracted, consumable interfaces to read from and write back to core systems themselves. Analytical integrations can serve to build out of the “enterprise data hub,” replicating data from core systems—often in streams or real time—for analytical use cases. Successful delivery of this foundational integration architecture layer requires a combination of strategic guidance from enterprise architecture to help steer teams on key design decisions. Additionally, new or underrepresented skillsets, such as data engineering, may be needed in the organization for it to scale, as well as a sound cloud strategy and infrastructure automation capability. Phase 2: Establishing consumable interfaces, integration points, and self-service tooling With the foundational integration patterns in place, the next step comprises the development of consumable interfaces and integration points, and the associated self-service tooling roll out. Socialization of the enterprise APIs, data marts, and available integration points through living documentation artifacts (like Swagger or wiki-hosted data catalogs) could open the doors for business as consuming apps, dashboards, automation bots, and other products begin leveraging the platform’s offerings. A self- service model is ideal, where consuming teams have everything they need at their fingertips to find, connect to, and communicate with points of integration across the platform. Sandbox environments (secure, isolated areas for experimentation with data and integrations) could be set up to encourage citizen development—a safe way to explore new use cases for harnessing the data within the platform. Phase 3: Incrementally scaling to additional domains and expand off-the-shelf accelerators Once the first end-to-end use cases are delivered for a given domain, phases 1 and 2’s processes can be repeated to expand the API and data catalogs with additional business domains, data sets, or system integrations. These could be based on use cases and prioritized by business value. In addition, the integration patterns defined in phase 1 could
development of the target state platform needs to be a collective effort—it cannot be achieved in isolation. Business stakeholders could consider partnering with the whole IT function to ensure the alignment of goals and outcomes, address dependencies, and reduce risk. Three phases of transformation We have observed that a successful transformation to a digital utility platform takes place over three phases. Phase 1: Developing foundational patterns for integration architecture The first step in implementing a target state platform model is to establish foundational patterns for system integration and platform consumption. The integration layer is a vital starting point—getting it right means fewer headaches during large system upgrades or consolidations in the future. Getting it wrong, however, can lead to multiyear overruns of large system modernization efforts. A sound integration architecture could decouple user-facing systems of engagement from the backing core IT systems, thereby reducing dependencies and eliminating sizeable tech debt. This often accrues when utilities build business functionality within rather than on top of core IT systems. With this layer of abstraction in place, future M&A efforts could be simplified and consolidations made easier. Applying the architecture principles from the lean governance model at the integration layer could serve as a replicable blueprint for scaling and expanding the platform over time. We recommend beginning by focusing on use cases within work, asset, or customer management as these core IT systems are central to many digital value cases. Organizations can start with one or two foundational use cases (such as customer payment journey or asset analytics) to prove end-to-end platform integration. For these use cases, the integration patterns can take two forms: operational integrations and analytics. Operational integrations can exist as managed enterprise
Accelerating the journey to net zero
136
Made with FlippingBook Online newsletter maker