Across mega-projects, new cities, destinations and infrastructure programs, organizations are investing in AI, digital platforms, procurement transformation, sustainability, advanced analytics and new operating models. Ambition is there.

But conversations around Saudi Mega Projects event we participated in a few weeks ago surfaced a harder question.

What Happens When All That Innovation Has to Work Together?

Many mega-projects are moving into, or have already reached, their operational phase. The challenges are consequently expanding into demand forecasting, operational planning, customer experience, resource planning and identifying partnerships capable of creating long-term value. At the same time, data remains distributed across stakeholders, systems and platforms.

That changes the intelligence challenge.

The next advantage may not come from adding another AI model, dashboard or point solution. It will come from connecting what already exists into an intelligent workflow capable of turning fragmented signals into coordinated action.

Intelligence Is the Glue That Connects Ambitions

Consider a decision that sounds relatively straightforward: selecting the right supplier.

In a mega-project environment, it rarely is.

The decision can touch procurement information, supplier intelligence, cost models, delivery milestones, geopolitical exposure, sustainability commitments, local-content objectives and operational requirements.

Now multiply that complexity across thousands of decisions.

One of the recurring issues we heard at Saudi Mega Projects event was precisely this fragmentation. Data quality, lack of context and integration remain significant hurdles because information sits across multiple stakeholders, platforms and systems. Supply-chain resilience is simultaneously under pressure, increasing the value of earlier visibility into potential shipping delays, capacity constraints, and other emerging risks.

The traditional response is often to solve each problem separately.

A procurement platform here.
A forecasting model there.
An AI use case somewhere else.
A sustainability dashboard alongside them.

Each may create value independently. But the bigger opportunity is connecting them.

This is why we believe the AI conversation needs to change.

Point solutions, however good, optimize only a certain area. For example, a risk alert only matters if someone can see what it means for cost, timeline and commitments.

Connection changes the math. When data, domain knowledge, analytics and AI work as one system, every decision draws on more context, and every outcome feeds back into the next one. Intelligence starts to compound instead of sitting in silos.

So the question for leaders is no longer:

Where else can we deploy AI?

It is:

How do we build an intelligence layer that connects data, domain knowledge, analytics and AI, so that every decision makes the next one better?

The Decision Chain

Imagine a project where a potential supplier disruption emerges.

In an integrated intelligence environment, that signal should not end with a risk alert.

It could become the beginning of a decision chain:

Signal → Context → Impact → Options → Decision → Action → Measurement

The disruption is identified. Relevant project and supplier context are brought together. Potential implications are assessed. Alternatives can be compared. Decision-makers receive evidence they can act on. And, critically, the result can be measured.

That final step matters.

Measuring value was one of the most discussed topics at the event. It starts with a framework that looks at value from several angles: revenue acceleration, efficiency, velocity, compliance. Then comes the harder part, designing how to measure it, which requires a deep understanding of how the processes actually work. And it cannot be set once and forgotten. As the project evolves, the measures should evolve with it. The payoff is a clear answer to one simple question: what is working, and what is not? That is a much more consequential question than how many AI pilots an organization has launched.

In our economic impact work, we see this play out every time. Recently, we presented an assessment of a multi-billion-dollar waterfront development to a national Economic Council. Ministers didn't ask about the model. They asked about the assumptions: how did you estimate tourist numbers? That is the real test of value. It has to stand up to scrutiny from the people who are accountable for it. This is why we combine agentic AI with domain economists.

The AI builds baselines and scenarios faster. The experts make sure every number can be defended. Speed without credibility doesn't help anyone make a decision.

Vision 2030 Raises the Bar from Project Delivery to Value Delivery

This matters particularly in the context of Vision 2030.

The challenge is not simply building ambitious projects. It is understanding and improving what those projects ultimately create.

That means looking beyond conventional measures of delivery into questions around economic growth, jobs, place, resilience, local content, tourism and sustainability. These are dimensions already incorporated into Evalueserve’s Economic Impact Assessment approach for infrastructure programs.

The intelligence model therefore has to connect decisions at several levels:

1. Project Intelligence

Are milestones, dependencies, risks and budgets pointing to emerging problems?

2. Commercial and Procurement Intelligence

Where are supplier, sourcing and cost decisions creating value or introducing exposure?

3. Market and Demand Intelligence

How will demand develop, and what does that mean for operators, capacity and commercial performance?

4. Economic Intelligence

What is the wider contribution of an investment to growth, people, place and resilience?

5. Sustainability Intelligence

Are commitments translating into measurable outcomes?

6. AI Value Realization

Where is AI genuinely improving efficiency, revenue, velocity or risk rather than simply increasing the technology footprint?

The opportunity is greater when those capabilities stop behaving like separate boxes.

The glue matters more than the pieces

This is perhaps the most important lesson.

Organizations Need a Way for Intelligence to Travel Across the Organization

That requires technology, certainly. But technology alone is insufficient. That combination is what we mean by the glue.

Evalueserve can sit between: data and context, technology and domain expertise,
insight and execution, AI adoption and measurable value.

The objective to shorten the distance between something changing somewhere in a complex ecosystem and someone making a better decision because of it.

We've seen what this looks like in practice. For a central Saudi ministry, we built an economic intelligence capability. It brings together secondary research, subscribed databases, expert input, and live dashboards and alerts in one place. The result was 35% faster time to insight and far fewer ad-hoc requests to the team. In tourism, we mapped and benchmarked 60+ operators and developed about 25 activation concepts for a Saudi region. That work was still being used for operator onboarding long after the engagement ended. Neither result came from a new platform. Both came from connecting what was already there.

What It Means in Practice

For mega-project leaders, we believe the next phase should begin with workflows, not technologies.

Take one high-value decision and map the intelligence around it.

For example:

A critical supplier signals a potential delivery problem. What needs to happen between that signal and an executive decision?

  • Identify which information is required.
  • Determine which systems contain it.
  • Add the external signals that could change the decision.
  • Bring in the domain expertise required to interpret them.

Use AI where it improves speed, synthesis or predictive capability.

Then measure whether the resulting workflow actually improved the business outcome.

That creates a very different AI agenda.

Instead of asking “Where can we use AI?”, leaders can ask:

  • Which decisions matter most?
  • What prevents us from making them earlier or better?
  • What intelligence is missing?
  • What should machines do, and where is human judgment essential?
  • How will we know the decision improved?

Our approach to AI transformation is simple: start with domain, scale with AI.

First, we ground AI in domain context, building in sector knowledge, organization-specific context and strong data foundations from day one, because AI without context produces answers, not decisions.

Next, we transform workflows with agentic AI, embedding agents with experts in the loop so that 15–30% efficiency gains can be reinvested in growth rather than simply banked. Finally, we scale what works. Shared pipelines, models and guardrails let results spread across the business, cutting total cost of ownership by 40–50%. Adoption is what turns isolated pilots into an enterprise capability.

At Saudi Mega Projects, almost every conversation came back to one point: leaders don't want more pilots, they want proof. Starting with the domain is how we provide it.

From a Portfolio of Innovations to an Intelligence System

Saudi Arabia has already demonstrated its capacity to pursue transformation at extraordinary scale.

As projects advance and operating complexity grows, competitive advantage will increasingly depend on how quickly organizations can connect signals across procurement, projects, markets, operations, sustainability and economic impact, then translate that intelligence into action.

That is the role we see for Evalueserve: not another disconnected piece of the technology landscape, but the intelligence layer that helps the pieces work together.

And that could prove crucial to the next chapter of Vision 2030.

Because ultimately, the measure of innovation is not how much of it you deploy.

It is how much better the Kingdom can decide because everything finally works together.

Our prediction: by 2029, the giga-projects that pull ahead won't be the ones with the most AI. They'll be the ones that treat demand and economic impact as live operating numbers, reviewed every month like cost and schedule. The rest will still be reporting what they built, when the real question will be what it is delivering. Alongside its mega-projects, the Kingdom is preparing to host Expo 2030 and the 2034 FIFA World Cup. We have no doubt that the build-out will be world-class. The harder question is what happens after the opening ceremony. Stadiums, hotels and destinations only succeed if people keep coming, not for a single event but every month of the year. That makes demand planning one of the most critical and most underestimated capabilities of the next phase. It means connecting visitor flows, event calendars, air capacity, pricing and experience into one view, long before the first guest arrives.

Without it, even the most impressive construction risks becoming capacity without customers.

Talk to the Authors

Dheeraj Gogna is the Vice President in Evalueserve’s Professional Services practice, with expertise in economic intelligence, market and strategic research, and decision support. His work helps organizations connect data, research, and domain insight to evaluate opportunities and understand economic and commercial impact.

Message Dheeraj on Linkedin

Michal Radziuk is a Director working with Evalueserve’s clients on data, analytics, and AI-led transformation, with a focus on turning fragmented information into connected intelligence and measurable business outcomes. His expertise spans AI-enabled workflows, decision intelligence, and applying technology within complex business environments.

Message Michal on Linkedin

Written By

Dheeraj Gogna

Vice President, Professional Services at Evalueserve •  Posts

Michal Radziuk

Director, Customer Success at Evalueserve •  Posts

Latest Posts