In a recent briefing with regional media, Dean Vaughan, Azul’s vice president for APAC, explained why Java remains central to AI growth – not just as a programming language for AI applications, but as the engine that drives the speed and scale of AI services.
AI – it mostly runs on Java
Dean spoke about Azul’s continued growth in Southeast Asia and in newer sectors like utilities, which he sees as driven primarily by the critical nature of national infrastructure. He explained, “They’re highly regulated and critical infrastructure organisations, so they must have licensed and supported software. But the real growth is due to AI.”
Reviewing the 2026 State of Java Survey & Report, Dean noted that when he talks to customers, he finds that much of the AI growth is coming from “all of the apps that sit behind it.”
A typical customer service application, for example, uses AI tools to automatically populate customer information and present prompts that help agents do their jobs. “All of that stuff is written into the application itself, and that is where this report focuses – on the (use case of) Java being used to add AI capabilities into these chatbot applications.”
The brain behind the responses
Dean also stressed that generative AI tools like ChatGPT, Claude, and others are front-end interfaces where users enter their queries. “The real brains lie behind these tools,” he said.
Components such as Kafka, HBase, and others are commonly used in big data and real-time analytics systems, where Java serves as the language used to build applications that interact with these components. The goal is to retrieve and process data in real time for AI-driven services like the customer service assistant mentioned above.
“We’re seeing huge growth in Kafka, Elasticsearch, HBase, Spark, and more – technologies that are required to quickly access the information the AI tool can generate responses from,” Dean said. “All of that needs to happen super fast, and all of that requires a huge amount of investment.”
Cloud-based costs
Dean added, “Even though Python is being used for the actual AI application, all the software that sits behind it to power that application is Java. So, Java is by far the most critical programming language when it comes to powering AI.”
For Dean, this is where Java’s strategic importance to AI truly lies – but it also raises other concerns.
He described a growing worry among senior decision makers at large payment processors he has spoken with. “Their biggest concern, consistently, is: ‘How do we manage the exploding size of our Kafka, Spark, HBase, or Elasticsearch environments?’ In many cases, these are running in the cloud, and they’re using cloud-based infrastructure to deliver services and meet the escalating demand for AI.
If they have optimised Java, they need less compute to deliver the same AI service, and we can help them save a lot of money in their cloud or infrastructure expenditure.
Dean Vaughan
“This is a problem for them because it is cost-restrictive,” he said, adding that for organisations looking to speed up their back-end infrastructure and power AI services, the Azul Platform Prime might hold the key.
Optimised Java as a cloud cost lever
According to the Azul survey, 97 percent of respondents have taken actions to reduce their public cloud costs, and 41 percent chose using a high-performance Java platform as one of their top five strategies.
Dean said, “If they have optimised Java, they need less compute to deliver the same AI service, and we can help them save a lot of money in their cloud or infrastructure expenditure.” He added that the high growth in Southeast Asia is being driven by the payment processing industry, thanks to a shift towards cashless transactions.
As a company that focuses 100 percent on Java, Azul has also made a strategic product acquisition in the form of Payara. Dean described the acquisition as bringing two complementary products: enterprise Java for traditional banks and legacy infrastructure, and a Java platform specifically built for Kubernetes and containers.
Speaking about the latter product, Payara Micro, Dean said, “Big tech companies love this Java because it is purpose-built,” and concluded, “There are other areas where we can expand our Java and add more value to our customers.”
