When 1+1=3: A Unified Decision Layer for Utilities

For utilities, the challenge is no longer whether they have enough data. They do. Customer systems, meter systems, billing platforms, asset applications, ERP, HCM, supply chain, finance, weather, demographics, regulatory inputs: the signals are everywhere.

The harder question is whether those signals can be brought together quickly enough, trusted enough, and acted on directly enough to change outcomes.

That is where the old equation breaks down. In theory, adding more data should create more intelligence. In practice, more data can create more friction. Every new source can introduce another integration, another reconciliation effort, another governance question, and another delay between insight and action. As that friction grows, the ability to realize value from the data shrinks.

For utilities struggling to connect data, insight, and action, a unified decision layer can provide a better path than adding more isolated dashboards.

What is a unified decision layer?
A unified decision layer connects operational, enterprise, and external data through shared definitions, governed models, analytics, and workflows to support coordinated decisions, actions, and insights.

When 1+1=3: A Unified Decision Layer for Utilities — *This AI-generated image was created by Oracle Utilities using ChatGPT by OpenAI.*

That is the idea behind connecting utility-specific operational intelligence with enterprise performance intelligence. One side brings together the core operational signals of the utility: customer, billing, meter, asset, grid, and engagement data. The other connects the back-office signals that shape enterprise performance: finance, supply chain, workforce, procurement, projects, and operations. Together, they can create something larger than either domain can deliver alone.

This is why 1+1=3.

The third value is not another system. It is the outcome that can emerge when operational and enterprise intelligence reinforce each other.

Turning Revenue Risk into Customer Action

Consider revenue risk. A utility may see arrears increasing, bad debt exposure rising, or cash flow pressure beginning to build. On its own, that financial signal is important, but incomplete. The organization still needs to know which customers are most at risk, what is driving the risk, which segments need intervention, what programs are available, and whether engagement is actually working.

When customer, billing, usage, income, energy burden, payment, engagement, and financial data come together, the conversation changes. A credit and collections leader can move from “arrears are rising” to “these customers are facing affordability stress, these accounts are most exposed, these geographies or segments need attention, and these outreach programs are most likely to help.”

That is a different kind of decision.

It is also a more human one. The goal is not simply to reduce bad debt. It is to intervene earlier, prevent avoidable hardship, improve program adoption, reduce billing contacts, strengthen customer satisfaction, and stabilize delinquency trends. Financial performance and customer outcomes become connected instead of competing priorities.

A Cross-Domain Intelligence Layer

The same pattern applies across the utility. Meter data can support usage insights, exception management, forecasting, and disaggregation. Asset data can inform maintenance costs, work orders, asset health, inventory, and resource planning. Grid insights can help identify transformer loading, EV detection, phase mismatches, and emerging operational risk. Back-office data can add financial exposure, procurement constraints, labor capacity, project spend, supplier risk, and enterprise performance context.

The value is not just that these data sets exist in one place. The value is that they are normalized into shared semantic models, governed definitions, and prebuilt metrics that business users can trust.

That matters because operational systems are built to run the business, not necessarily to explain the business. A billing system, meter data management system, asset system, or ERP application may be excellent at processing transactions. But analytics requires a different structure: business-friendly definitions, consistent KPIs, flattened relationships, reusable metrics, and the ability to connect domains without every question becoming a custom data engineering project.

This is where prebuilt intelligence can change the economics of analytics. Instead of starting every initiative by moving data from production systems to staging areas, operational data stores, data lakes, and custom reporting layers, utilities can start from productized pipelines, shared data models, and ready-to-use content. That does not eliminate the need for strategy or expertise. It redirects expertise toward higher-value work.

The people closest to the business should be spending less time hunting for data and more time deciding what to do.

AI Raises the Stakes

AI raises the stakes further, but only if the foundation is right. AI without trusted data and clear business process can be just another source of noise. For AI to support utility decision making, it needs good data, business context, and compute. To influence outcomes, it also needs to appear where decisions actually happen.

AI-enabled decision-making can take three practical forms.

First, ask. Natural language experiences allow users to query governed data without needing to understand the schema, build a report, or wait for a specialist. A manager can ask for project counts, budget comparisons, overdue balances, risk segments, or trends, and receive visualizations and narrative explanations grounded in trusted enterprise data.

Second, act. Intelligence should be embedded into workflows, not bolted on beside them. If analytics identifies a collection risk, a meter exception, a supply disruption, a project cost anomaly, or a customer segment needing assistance, the next step should be close at hand. The user should not have to swivel between disconnected systems to turn insight into action.

Third, extend. With proper governance, agentic AI creates the possibility of reusable patterns that monitor conditions, interpret enterprise context, and recommend or initiate next steps. For example, a regulatory or supply chain change could trigger analysis of transformer inventory, asset risk, and procurement timing, helping the utility act before a constraint becomes a crisis.

The Multiplier Effect

This is the real multiplier. The first wave of analytics helped organizations see more. The next wave must help them decide faster and act with greater confidence.

For utilities, that shift is essential. The industry is managing challenges such as affordability pressure, grid modernization, electrification, distributed energy growth, aging infrastructure, workforce constraints, regulatory scrutiny, and evolving customer expectations. None of these challenges live neatly inside one application or one department.

A customer affordability issue is also a revenue issue. A transformer risk may become a supply chain issue. A field operations constraint may become a customer experience issue. A capital project delay may become a financial planning issue. The business is interconnected, so the intelligence layer must be interconnected too.

That is why connecting utility operations and enterprise performance is more than an analytics architecture. It is a way to simplify the equation utilities face every day.

Bring together the core utility signals. Bring together the enterprise signals. Reduce the friction between data, insight, and action. Then apply AI in ways that help people ask better questions, act sooner, and extend intelligence across the business.

That is how 1+1 becomes 3.

Not because the math changed, but because the outcome did.

About this Article

Generative AI was used to compile and summarize content presented by the author during an educational session at the Oracle Customer Edge Summit.

About the Author

Kojo Quaye, PE, is a Principal in Product Strategy for Data & Analytics at Oracle Utilities, where he helps shape the commercial and product strategy for data, analytics, and AI offerings serving the utility industry. He works at the intersection of utility business needs, customer data, product strategy, and go-to-market execution —turning market signals and customer priorities into practical ways for utilities to use data and AI to improve customer experiences, accelerate energy programs, and make better operational and commercial decisions.

Over more than 12 years in energy, consulting, and B2B software, Kojo has built a career around translating technically complex problems into clear, evidence-based decisions. At Oracle, he leads strategy for Oracle Utilities Data Intelligence and related analytics and AI offerings, partnering with product, sales, engineering, pricing, marketing, and utility customers on market segmentation, buyer needs, packaging and pricing, and field enablement. His work has included helping scale access to Opower insights by advising on deal economics, usage-based analytics pricing, value propositions; and helping teams connect product investment to customer outcomes and commercial opportunity.

Kojo is particularly interested in the conditions that make analytics and AI useful in the real world: trustworthy data, clear business context, measurable value, and a practical path from insight to action. He has contributed to Oracle Utilities thought leadership on customer engagement and energy efficiency, and regularly speaks with utility leaders and Oracle teams about data intelligence, customer analytics, and the evolving role of AI in the energy transition. Before Oracle, he was a Senior Consultant at Guidehouse, supporting demand-side management program evaluation, and an engineer at Enbridge Energy. At Enbridge, he worked across control systems, project engineering, and facilities integrity ensuring energy could safely be delivered.

Kojo holds an M.S. in Energy Science, Technology, and Policy from Carnegie Mellon University and a B.S.E. in Chemical and Biological Engineering from Princeton University. He is a licensed Professional Engineer in Mechanical Engineering – Thermal and Fluid Systems.