29-07-2025
Ecosystm's Tim Sheedy highlights ANZ's AI adoption challenges at Neo4j GraphSummit
At the Neo4j GraphSummit in Sydney, Tim Sheedy, Vice President of Research and Chief AI Advisor at Ecosystm, shared his insights on the state of Generative AI (Gen AI) adoption across Australia and New Zealand, highlighting both the potential and challenges the region faces.
"The vast majority of organisations are somewhere in the middle," Sheedy admitted to the audience. "Some organisations are deploying within business units. Some have gone out across their entire organisation. But just to let you know, the vast majority of them - they're not AI-first companies."
Sheedy observed that most companies are still in a "consolidation phase," focused on setting up the data pipelines and AI infrastructure needed for transformation, but not yet seeing AI-driven organisational change.
"76% of firms say they are consolidating, meaning they're getting their data ready, developing AI strategies, and piloting AI applications, but they don't see themselves as fully transforming with AI yet."
Despite these challenges, Sheedy highlighted some successful use cases of AI adoption, particularly in customer experience.
"Organisations that did it well reported a 52% improvement in their customer experience scores," he said. But he emphasised that AI is primarily being used to improve operational efficiency, with businesses seeking ways to do more with less.
"Business leaders want to do more with less or more with what they've got today. They all see AI as a fantastic opportunity to drive productivity."
Sheedy pointed to a growing trend of AI applications in operational tasks such as intelligent document processing, inventory management, and code generation.
"More than half of organisations are already using Gen AI for these tasks, and about 60% are using it in their help desks," he explained. However, he clarified that AI is still not widespread across all functions within most businesses.
"They're not saying all of their document processing across their organisation is intelligent, but they are using it somewhere in their business."
Looking ahead, Sheedy observed the rapid growth of AI usage.
"AI usage is growing, and it's growing quickly," he said. "In addition, areas like cloud resource allocation, software development, and fraud detection are set to experience further AI intervention." He also shared an example of a multinational company that significantly reduced proposal turnaround times, from six weeks to just two days, using Gen AI. "By using Gen AI, they're massively reducing that time to getting that first draft of the proposal together," Sheedy added.
However, Sheedy cautioned about the risks of hasty AI adoption.
"72% of consumers avoid brands after AI errors. And I'd say this is the same as employees in your organisation," he warned.
"You have a bad experience with a chatbot, you're probably not going to use that chatbot again in six months' time, even though it might be much better."
He also referenced the infamous Zillow case, where an AI-driven real estate algorithm lost USD $500 million due to flawed predictions.
"AI can have significant financial consequences if it goes wrong," Sheedy said.
Regulatory and organisational challenges are also slowing adoption. Sheedy highlighted that while Australian regulators have been more consultative, the costs of getting things wrong are significant. "It really is important to get Gen AI right the first time so you don't lose that trust, you don't lose that money, and you don't lose that time," he added.
Skills shortages and data challenges are the primary barriers to AI adoption in ANZ.
"The skills shortages and the data challenges are far more persistent and tougher to address," Sheedy explained. "At the moment, there are thousands of organisations in Australia trying to hire people with skills that either don't exist, or if they do exist, they're not going to get them for what they're paying."
Sheedy also emphasised the need for better data management. "Our data isn't clean enough. Our data doesn't understand context," he said. He advocated for a new approach to data, particularly through the use of graph databases, which provide a "semantic AI" layer that enables businesses to understand the relationships between data points.
"Graph database gives your data within your organisation the same capability as the data that sits within those language models. It helps your organisation understand the relationships between data - it's that semantic relationship understanding piece that you don't have," Sheedy explained.
Sheedy also pointed out that AI adoption is often hindered by the difficulty of changing core business processes.
"You have to change a process within your organisation in order to make that AI do something better or different," he said. This challenge is particularly evident when organisations need to integrate AI with legacy systems.
At the organisational level, Sheedy highlighted the need for better understanding of AI.
"People in our organisation don't understand our companies. They don't understand what AI can and cannot do. We need to actually upskill broadly across the business," he said.
Drawing from his experience with cloud computing, he suggested that AI adoption will only accelerate once organisations train their entire workforce on the technology's capabilities and limitations.
"I believe that AI will go into hyper growth when organisations are starting to train up all of your employees across the entire business on what AI is, what it can do, what it can't do, starting with the board of directors and the CEO."
Sheedy also stressed the importance of embracing emerging technologies like graph databases to stay competitive.
"Graph DB is going to be everywhere within our organisations," he said. "We're at the beginning of a massive wave, and it's going to reshape the web and the way we understand data. We need to get on board early and help drive the transformation."
