c-suite

AI as an enterprise performance lever in the energy transition

insight
July 06, 2026
9 min read

Authors

Thomas Steirer

 

Thomas Steirer is Chief Technology Officer (CTO) at Nagarro. His focus is on developing scalable and sustainable solutions that are primarily designed to deliver valuable information.

 

 

 

Eugen Rosenfeld-modified


Eugen Rosenfeld is a CTO & a Solution Architect in Life Sciences at Nagarro. He has more than 20 years experience in different programming languages, technologies and business domains.

Vision got us here.

The energy transition is entering a phase where execution excellence will determine which companies emerge as leaders. The first decade was defined by vision, capital commitment, and policy alignment. The next phase will be won by companies that execute better - not simply those with the boldest strategies - while managing volatile commodity markets, escalating regulatory pressure and increasingly complex operations across the full value chain. 

The companies that win will not simply be those with the best strategies on paper - they will be those who can move faster, decide better, and operate more efficiently than their peers. The competitive battleground has shifted from strategy to execution, and artificial intelligence is becoming the execution engine behind that advantage. How AI is positioned, governed and measured - as an enterprise performance lever rather than an IT initiative - will determine whether it compounds into real advantage or accumulates as expensive technical debt. 

18–25%
Reduction in maintenance costs achieved through AI-enabled predictive maintenance.
(source: Mckinsey) 
Up to 50%
Reduction in unplanned downtime through predictive maintenance.
(source: Mckinsey) 
20–40%
Increase in maintenance productivity using GenAI-enabled maintenance workflows.
(source: Mckinsey) 

Where your rivals are already placing their AI bets

European energy majors and regional players are each taking a distinct posture toward AI, and their recent moves point to the same conclusion: the competitive battleground has shifted from strategy to execution. The leaders are no longer those with the boldest transition pledges, but those turning AI into measurable operating performance across refining, trading, retail and reliability

Company

Strategic posture

AI & digital priorities

Signal

Shell
Performance & simplification
Shell.ai scaled across thousands of assets; AI trading desks and predictive maintenance
Scale · Trading AI
BP
Returns focus, capital discipline
Palantir-powered operations; predictive maintenance and emissions analytics tied to capital returns
Reliability · Efficiency
TotalEnergies
AI industrialization at scale
Generative-AI rollout with Microsoft; industrialized data products and LLM-powered operations
Foundation · Platform
MOL / ORLEN
Central European consolidation
Digital refinery programs, retail app ecosystems and grid integration across Central Europe
Regional · Growth
Integrated energy leaders
Integrated transition leadership
Fluidic Intelligence across the full value chain - refining, chemicals, retail, EV, ReOil, gas, circularity
Integrated · Differentiated
Leading integrated energy companies demonstrate a genuinely distinctive position: integrated assets spanning conventional and transitional energy, circular economy, and new mobility, combined with a growing AI foundation. These companies show that the opportunity is not to imitate peers, but to turn integrated breadth into a multiplier for connected intelligence.

Connected intelligence for a frictionless enterprise

In an integrated energy portfolio, the greatest single drag on performance is rarely a lack of data, ambition or talent - it is friction. Friction between systems that do not speak to each other. Friction in decisions that take longer than the market allows. Friction in operations where complexity compounds small inefficiencies into significant cost. Friction in regulatory processes where reporting lags operational reality by weeks or months.

The term fluidic is deliberate. Fluidic systems are adaptive, connected, and efficient under pressure. They find the path of least resistance without losing force or direction. Applied to an enterprise as complex as an integrated energy company, Fluidic Intelligence describes AI that reduces operational resistance, enabling capital to flow to its highest-value uses, decisions to reach their optimal quality faster, and assets to perform closer to their designed potential.



The six dimensions of  enterprise performance
— enabled by fluidic intelligence 

Capital discipline

Portfolio · Allocation

Margin protection

Refinery · Retail · Trading

Asset reliability

Predictive · Prescriptive

Regulatory confidence

Compliance · Reporting

Transition execution

ReOil · EV · Circularity

Better execution

Speed · Quality · Agility

This moves beyond AI as a technology initiative, positioning it instead as an enterprise performance philosophy, embedded deeply inti each decision, every operation, and all the risks the business manages. The shift required is from operational complexity to operational clarity: the integrated portfolio creates both the need and the opportunity for this kind of connected intelligence at scale.
Fluidic Intelligence is connected intelligence that helps the business move with more precision in markets that are becoming more volatile, more regulated and more unforgiving.

Thousands of better decisions, every single day

Connected intelligence earns its value where decisions are made. Across an integrated portfolio, the same foundation improves judgment in every business, turning fragmented data into faster, better-informed action.

Refining operations


Real-time guidance on crude selection, unit scheduling and yield trade-offs lets planners capture margins that manual cycles leave on the table.

Retail & EV charging


Demand-aware pricing, site-level assortment and charger placement sharpen forecourt returns and keep EV networks utilized as mobility patterns shift.

Chemicals production

Process optimization and quality prediction lift throughput and energy efficiency while reducing off-spec batches and unplanned slowdowns.

Trading & forecasting

Sharper demand, supply and price forecasts give trading and gas desks earlier, more confident positions in volatile markets.

Circularity (ReOil)


Feedstock matching and yield optimization make pyrolysis and recycling routes such as ReOil more reliable and economically scalable.

Compliance & reporting

Automated data lineage and emissions reporting compress regulatory cycles from weeks to days and keep audit trails always ready.

Five principles that turn pilots into performance

Most AI initiatives in large energy companies follow a familiar arc: promising pilots, enthusiastic early results, and then a gradual plateau as scale proves elusive, governance is unclear, or business adoption stalls. These five principles are designed to break that pattern - ensuring that AI investment compounds in terms of value over time rather than accumulating as technical debt.

01 start from business friction, not technology trends

title-underline

The most durable AI investments solve real, measurable business problems - not technology aspirations. Leading companies orient their AI portfolio around areas where fragmented data, manual decision-making, operational complexity, or slow execution create quantifiable cost, risk, or delay. The question is always the same: where does the business move too slowly, and what does that cost? Friction is the starting point; AI enables the flow.

Focus areas: Refinery margin optimization · Predictive maintenance · Regulatory reporting · Demand forecasting
02 build a focused AI value portfolio

title-underline

Not all AI use cases are equal, and resource diffusion is the most common reason enterprise AI programmes underperform. Leading operators maintain a prioritized, governed portfolio of AI investments - assessed on business value, feasibility, risk, and scalability. Isolated pilots that cannot scale create documentation, not enterprise value. Priority goes to use cases that run across multiple business units or replicate across markets without rebuilding from scratch.


Portfolio criteria: Business value · Feasibility · Risk · Scalability · Replicability
03 strengthen decision support before automation

title-underline

In regulated and safety-critical operating environments, the instinct to automate must be tempered by a commitment to traceability, human accountability, and decision quality. The correct sequencing is to first build AI that makes human decisions faster and better, then automate where confidence and governance are established. Premature automation in refining, emissions compliance, or grid management creates risk that outweighs the gains. Recommendation systems build trust; automation harvests it.

Priority: Recommendation layers · Audit trails · Human-in-the-loop design · Regulatory traceability
04 scale through reusable foundations

title-underline

Every one-off AI solution creates two costs: the cost of building it, and the hidden cost of maintaining it in isolation. An AI program at enterprise scale only achieves enterprise-grade scale when it is built on shared foundations - data products that serve multiple use cases, governance patterns applied consistently, integration layers that connect AI outputs to operational systems, and AI agents that can be retrained and redeployed rather than rebuilt. Platform thinking is the difference between a portfolio worth far more than the sum of its pilots and one worth merely the sum of its pilots.


Build: Data products · Governance patterns · Integration layers · AI agents · Platform capabilities
05 measure business outcomes, not technical metrics

title-underline

The performance of an AI program should be reported in the language of the business, not the language of data science. Model accuracy, training speed, and inference latency are engineering concerns - not executive ones. The metrics that matter to leadership are: avoided downtime expressed as cost or production; energy efficiency gains measured in tons of CO2 or euros; forecast accuracy measured in margin impact; compliance traceability measured in audit readiness; and execution quality measured in speed-to-market. Measured this way - not by pilot counts or model accuracy - AI earns its place alongside capital, talent and strategy.

Measure: Avoided downtime · Energy efficiency · Forecast accuracy · Compliance readiness · Execution speed

Stop treating AI as an IT project

The most important reframe in this document is also the simplest: AI must be positioned and governed as an enterprise performance lever, instead of an IT initiative, a digital transformation program, or a collection of technology pilots. When AI is owned by technology, it optimizes for technology. When it is owned by the business, it optimizes for business performance.

icon     Margin Protection
Refinery · Retail · Trading 

icon      Asset reliability
Predictive · Prescriptive 

icon      Capital discipline 
Portfolio · Allocation 

icon      Regulatory confidence
Compliance · Reporting 

icon      Transition execution
ReOil · EV · Circularity 

 

Quality engineering leader-strategy
An integrated asset base - spanning conventional and transitional energy, chemicals, circular economy, and new mobility - creates a rare opportunity for AI that connects performance across the whole enterprise rather than optimizing isolated silos. This is the integrated advantage: not just AI for its own sake, but Fluidic Intelligence as a coherent strategy for performing better in a world that is becoming more volatile, more regulated, and more complex by the year.

Winning the transition through execution.

The energy transition will not be won in the boardroom or the policy forum. It will be won in the refinery control room, the retail forecourt, the carbon registry, and the capital allocation meeting - through thousands of better decisions, made faster and with more confidence.

Fluidic Intelligence describes how that happens - not as a single initiative, but as a continuously improving enterprise capability that reduces friction, amplifies performance, and creates the confidence needed to navigate the transition successfully. The five principles are not a checklist; they are a discipline - one that, applied consistently, clearly separates the companies that execute from those still treating AI as an experiment.

The companies that will lead the next decade of energy are not those who adopted AI the earliest - they are those who embedded it most deliberately into how they make decisions, operate assets, and execute their transition strategy.

The five-point leadership agenda

The companies pulling ahead treat Fluidic Intelligence as an enterprise performance lever. Five priorities mark how they put it to work - and leading integrated energy companies illustrate each in practice.

Identify the friction


They map the highest-value friction points across the value chain where AI can create measurable impact within twelve months - starting from operations, not from the technology catalogue.

Build the portfolio


They run a governed AI value portfolio with explicit criteria for prioritization, clear progression gates, and disciplined sunset conditions for initiatives that do not scale.

Invest in foundations

They accelerate shared data products, integration layers, and AI platform capabilities that serve the whole portfolio - not individual projects.

Define the metrics

They measure AI through business outcomes - margin protection, reliability, compliance confidence, and execution speed - not pilot counts or model metrics.

Govern at the top

They place AI governance at the C-suite level, keeping accountability firmly with the business - not with technology. Fluidic Intelligence is a leadership agenda, not an IT agenda.

Leading integrated energy companies - with integrated assets, ambition, and a growing technical foundation - show what it takes to lead the AI-enabled energy transition. The future advantage in energy will come from decision quality and execution speed, with intelligence flowing seamlessly across the enterprise. The next phase belongs to those who execute better.
Get in touch

Build your AI advantage. Lead with connected intelligence.