PKN EXCLUSIVE: Melbourne-based technology company Phantm has just launched Phantm AI, an image-based tool designed to turn packaging visuals into structured data that helps FMCG brands identify cost and waste reduction opportunities. PKN speaks to Ed Whitehead, head of strategy at Phantm.
As Australian brands face increasing pressure to reduce waste and prepare for tighter packaging regulations, access to reliable data is emerging as a key enabler of meaningful change. Yet, for many businesses, capturing and managing packaging data at scale remains a complex task.
Phantm has set out to tackle this challenge with the launch of Phantm AI, an image-based technology that converts packaging visuals and documents into structured, usable data – giving FMCG companies a fast, accessible way to identify cost and waste reduction opportunities.
According to Ed Whitehead, Phantm’s head of strategy, the tool was developed to make engagement easy and accelerate the first step in data capture.
“The main drive behind developing the image-based tool is ease of engagement – it’s about making it easy for customers to get started,” Whitehead explained. “Ultimately, one by one, image recognition isn’t the end goal. The technology is built to service bulk data ingestion. But images are the entry point – they help businesses see how quickly packaging can be turned into data, and that’s what builds confidence.”
From image to insight
Phantm AI uses machine learning to interpret product images or URLs, identifying indicative materials and weights, and then placing this information within Phantm’s queryable database. The platform ingests data from multiple sources – including PDFs, spreadsheets, and procurement systems – creating a structured, shareable packaging library that can be used for reporting, analysis, and compliance.
Whitehead likens this shift to how accounting software transformed business operations.
“People are familiar with data. They trust it. Think about how accounting changed with Xero – it made data usable and visible. We’re doing the same for packaging,” he said. “Our goal is to turn packaging into data really quickly, then refine it over time.”
Bridging human expertise and AI
While artificial intelligence drives Phantm AI’s automation, Whitehead emphasised that human validation remains central to ensuring accuracy and value.
“Data quality is what matters – and that requires humans to teach the technology,” he said. “AI and machine learning on their own are pretty useless without human integration. It’s people who provide the logic – what’s the thing, what’s its mass, what’s its purpose, how do we optimise it?”
He added that Phantm’s workflows are designed to interrogate and refine “scrappy data” from customers, transforming it into structured datasets that hold long-term business value.
“Completing an APCO report is a low-value application of data,” Whitehead said. “The real value comes from optimisation – the insights that help businesses save money, simplify materials, and move faster.”
Driving visibility across the business
Phantm AI is designed to support collaboration across departments – from sustainability and operations to procurement and finance – by making packaging data visible and meaningful to all.
“Visibility aids the conversation,” Whitehead said. “Not everyone cares about the data itself, but they care about what the data shows. When you tell a CFO or a legal team that a packaging format represents 20 tonnes of material, that has weight. It prompts conversations about efficiency and impact.”
He noted that the platform enables companies to identify problematic materials, manage risk, and create efficiencies, with reporting that can be repurposed for multiple uses – from APCO compliance to lifecycle assessments and digital product passports.
Preparing for a regulated future
With Australia expected to strengthen packaging and waste regulations in the coming years, Phantm believes its platform can help brands prepare early. Drawing on his experience in the UK, Whitehead said that regulation has proven to be an important driver of data adoption.
“In the UK, companies are now required to consolidate and report their packaging data – and the government even directs them to platforms that can help,” he said. “It all comes down to evidence and audit. Phantm’s goal is to become a trusted source of that data.”
While compliance is a natural outcome of accurate data, Whitehead said it is not Phantm’s primary focus.
“Compliance is important, but it’s a by-product of good data,” he said. “Our focus is helping businesses save money through material optimisation and simplification. When you do that well, compliance follows.”
Opportunities in material reduction
Whitehead pointed to secondary packaging – particularly fibre-based formats – as an area ripe for improvement.
“Secondary packaging makes up around 50 per cent of most product portfolios, and there’s a lot of optimisation that can be done there,” he said. “It’s often an undervalued cost in the supply chain, but our data brings those opportunities to light very quickly.”
Catalyst for digital adoption
Whitehead believes that widespread adoption of digital packaging intelligence in Australia will ultimately be driven by regulation – but that early movers stand to benefit now.
“We need to digitise this information – it’s infinitely more powerful when it’s structured and shareable,” he said. “Most businesses don’t sell packaging; they buy it as part of a product. Data helps them understand what they’re buying, why they’re buying it, and where they can improve.”
He added that Phantm AI enables procurement teams to see opportunities they may not have had time or visibility to address, often revealing unexpected efficiencies.
AI’s broader role in packaging sustainability
Looking ahead, Whitehead sees enormous potential for AI and machine learning to improve packaging sustainability across the value chain.
“If we can map materials at the front end, we can feed that data into downstream systems,” he said. “It’s a much smarter way to capture information than trying to sort it at the end of life. AI gives us the ability to identify, categorise, and track materials in a way that supports circularity.”
He pointed to emerging technologies in recycling and material identification – such as those being developed for soft plastics recovery – as examples of how AI could link packaging design and waste management in the future.
“In the end, there aren’t that many packaging formats out there,” he said. “Once we’ve mapped them properly, the data becomes incredibly powerful.”
Phantm’s new AI technology represents a step forward in connecting packaging, data, and sustainability. By simplifying how packaging data is captured, validated, and used, Phantm aims to help Australian companies move beyond compliance towards proactive, cost-effective impact reduction – one data point at a time.
