Inside AZX: Why AI in critical industries only works when it’s embedded in reality

AI adoption is accelerating at a remarkable pace. New models, new tools, new benchmarks appear almost weekly. And yet, in critical industries - energy, utilities, infrastructure, and real estate - progress often feels slower, more cautious, more uneven.

According to Aaron Goldfeder, Founder and CEO of AZX, that gap isn’t a failure of ambition or investment. It’s a reflection of reality.

“These industries don’t get to experiment in a vacuum,” Aaron explains. “They operate under regulation, legacy systems, and real-world consequences. When something breaks, it actually matters.”

That reality is what sits at the core of AZX.

The Interview: Inside AZX

Q: What “aha” moment made you realise this was the right problem to solve?

Aaron traces the origin of AZX back to a moment of sharp contrast. He was attending a Climate Change AI event at NeurIPS in 2019, surrounded by some of the brightest minds in artificial intelligence. The room was packed with researchers applying AI to climate and energy challenges. The ideas were ambitious. The models were impressive.

But something felt off.

Having spent nearly a decade working directly with utilities, Aaron couldn’t ignore how disconnected many of the conversations were from operational reality. “It was obvious,” he recalls. “Most of the AI people didn’t really understand how these industries function day to day. And at the same time, the people running those industries didn’t yet understand what AI could actually enable.”

The gap was striking  and enormous. “That day made the problem very clear,” Aaron says. “But I wasn’t quite ready to tackle it yet.”

It was only after his previous AI company was acquired that everything aligned. The tools had matured. The timing was right. And the gap he had observed years earlier was still wide open.

Q: Can you share a recent example of real-world impact?

When asked about tangible impact, Aaron doesn’t point to abstract metrics or model performance. Instead, he talks about customers - and their customers.

One recent engagement helped a client unlock meaningful progress with their own end users. The result wasn’t just operational improvement or marginal optimisation. It led to sharp revenue growth alongside positive climate outcomes.

“That’s the kind of work that’s most satisfying,” Aaron says. “So much of AI is framed around cutting costs. But growing the pie - especially in these industries - changes the conversation entirely.”

It’s also where adoption accelerates.

Q: What is the boldest risk you have taken so far - and how did it pay off?

Aaron doesn’t hesitate. Starting AZX fully bootstrapped. At the time, it felt risky. In hindsight, he views it as foundational. Bootstrapping gave the team space — space to define their culture, their standards, and what they were willing to stand for.

“There was no outside pressure,” he explains. “Only our customers, partners, and employees.” That constraint also imposed discipline. Every engagement had to deliver real value. Every deployment had to work. That focus on execution and trust remains central to AZX today.

‍Q: What  lesson has surprised you most when working with traditional industries?

Aaron is quick to emphasise trust. Leaders in critical industries are constantly approached by vendors promising transformation. Over time, that breeds skepticism — often for good reason.

“Being trustworthy and keeping things real is sadly a huge differentiator,” he says. That trust is built not just by delivering results, but by being honest when things don’t go perfectly — and by occasionally telling hard truths about an organisation’s own processes. When done with candor and respect, Aaron has found those conversations are often deeply valued.

‍Q: What are the long-term ripple effects if AZX succeeds?

AI and analytics spending in critical industries is accelerating rapidly, often reaching hundreds of millions of dollars per organisation. Yet much of that spend still fails to deliver durable change.

Aaron believes that’s because the wrong delivery models are being applied. The next generation of AI companies in these sectors, he argues, won’t look like traditional SaaS vendors — or traditional consulting firms. They will be built by operators. They will embed alongside customers. And they will design systems that respect operational, regulatory, and human realities from day one.

That’s the future AZX is building toward: AI systems that strengthen decision-making, build trust, and enable progress in the industries that underpin everyday life.

About AZX

AZX builds forward-deployed, services-led AI platforms for critical, asset-heavy industries including energy, utilities, infrastructure, and real estate.

Founded by seasoned operators with deep domain expertise, AZX embeds alongside customers to design AI systems that integrate securely into existing environments — addressing unstructured data, fragmented workflows, and regulatory complexity.

By pairing embedded delivery with reusable platform technology, AZX enables AI systems that scale trust, adoption, and real-world impact.
‍

Continue reading more about AZX and why we invested or visit their website at www.azx.io

---

Note. article written by Pauline Jimenez, Head of Marketing at KOMPAS VC

‍