Everyone is talking about AI, but many organisations are still asking the same question: Where do we actually begin?
From generative AI and copilots through to autonomous agents and computer vision, the opportunities are growing rapidly. Yet for many Australian organisations, moving beyond experimentation and into production remains a significant challenge.
We sat down with Marlon De Leon, AI Solutions Architect at HPE, to discuss what he’s seeing across the Australian market, why so many AI projects stall before they deliver value, and how organisations can build the right foundations from day one.
“I’d say most organisations are still in the infancy stage of their AI journey.”
“Many businesses are already using AI in some form, whether that’s Microsoft Copilot, ChatGPT, Gemini or other AI-powered tools. But using AI assistants isn’t the same as embedding AI into your business.”
“What we’re seeing is plenty of curiosity and experimentation, but relatively few organisations have moved into production at scale.”
There are encouraging signs across industries including Agriculture, logistics, Financials, Retail and Manufacturing, where organisations are beginning to explore applications such as computer vision and digital twins. However, compared with some overseas markets, Australia is still at the beginning of its enterprise AI journey.
“Everyone wants to talk about platforms, infrastructure and architecture.”
“But before any of that, you need to define the problem you’re actually trying to solve.”
According to Marlon, this is where many AI projects lose momentum.
Rather than asking Which AI platform should we buy?, organisations should first ask:
“The technology comes afterwards. If you don’t define the problem first, it’s very difficult to know whether your AI project has actually been successful.”
“I sometimes refer to it as the ‘proof of concept graveyard’.”
Many organisations successfully build an AI proof of concept, but very few successfully scale it across the business.
Marlon believes there are several reasons for this:
“When organisations try to scale without those foundations, that’s where projects often stall.”
“Data is probably the single biggest factor in whether an AI project succeeds.”
“It’s not simply about having lots of data. It’s about having the right data for the problem you’re trying to solve.”
Preparing data includes understanding where it resides, ensuring it is accurate, applying the appropriate governance controls and determining how sensitive information should be managed.
This is also where sovereignty and security become increasingly important.
Many organisations are now looking at hybrid AI approaches that allow them to deploy AI closer to where their data already resides, rather than moving sensitive information unnecessarily.
Solutions such as HPE Private Cloud AI, HPE GreenLake and HPE AI Factory – Sovereign AI support this approach by enabling organisations to modernise AI while maintaining control over data, governance and compliance requirements.
“My advice is simple. Don’t start with the technology.”
“Start by defining the problem.”
Once the business challenge has been identified, organisations can begin evaluating the data they have available, the governance required and the architecture needed to support long-term success.
“This is where working with experienced partners can make a real difference. The right conversations early in the process help organisations avoid costly detours later.”
Through their partnership, Astrontech and HPE help organisations assess AI readiness, identify practical use cases and build a roadmap that takes AI from experimentation to production.
Every organisation’s AI journey is different, but one thing remains consistent: successful AI projects begin with strong foundations.
If you’re exploring how AI could support your organisation, Astrontech and HPE are offering a complimentary ICT Strategic Review and AI Readiness Assessment to help you:
Book your complimentary ICT Strategic Review below today.
