Exclusive: Nuix CEO warns 'poor data in AI only results in poor decisions faster'

AI is predicted to have a bigger impact on our lives than the internet and smart phones combined.

For Nuix CEO Jonathan Rubinsztein, he agrees this is an evolutionary moment for business. "It's AI or die," he said. "Those who don't adapt, won't survive."

Nuix, he argued, is uniquely positioned to help businesses make that adaptation responsibly. "We've been working with data for 25 years," he said. "We understand it at the binary level and we are the tool of choice for nearly 100 of the largest regulators globally".

Unstructured, fragmented, and siloed data represents a persistent and formidable challenge for businesses, defying traditional tools and methods. Nuix's newly launched AI Data Curator is positioning itself as "best-in-class" solution to overcome these barriers. 

Unstructured data makes up over 90% of all enterprise data and is the fuel that drives GenAI. Well-curated data is the foundation for determining the success or failure of an organization's GenAI program. Even the most sophisticated AI-enabled system will fail to deliver reliable, explainable, repeatable, and scalable business value if the data that feeds it is inaccurate, irrelevant or poor-quality.

That's where Nuix's AI Data Curator comes in - a tool designed to help organisations clean, enrich and organise unstructured data, such as emails and document repositories, before training AI models.

"The success or failure of your Gen AI project is all about the data," Rubinsztein said. "Poor data in means poor decisions out - only faster."

Nuix's AI Data Curator has become so critical to this process that Rubinsztein now calls it "the Ozempic of the AI revolution".

"If you train AI on everything, it understands nothing," he explained. "Like Ozempic, it slims things down - trims the fat. We help you find the right data and feed it the right information."

That approach, he said, stands in contrast to companies who install off-the-shelf Gen AI tools but fail to structure their internal data in a meaningful way.

"Many organisations are buying Gen AI but struggling to execute," he said. "They've bought the tool, but they can't use it. And that becomes a failed project."

That failure, he warned, carries strategic risk. "If you're not accessing AI, your competitors probably are. And they're adapting faster than you are."

The company has also invested in building explainable AI frameworks, including Nuix Neo, its proprietary deep learning engine.

"We show our customers exactly how we built our models," he said. "It's completely transparent and ethical."

That transparency extends to how clients can choose to use AI within the Nuix platform. "We offer three models - Nuix-owned AI, Nuix-recommended third-party AI, and bring-your-own AI," he said. "And for each, we give you a methodology and structure to use it responsibly."

Crucially, Nuix has made its tools accessible to users beyond data scientists. "You can tune our AI like a dial," he said. "You don't need a PhD - you just choose the level of accuracy you want."

To that end, Nuix has developed clear ways to distinguish between synthetic and original data. "In our user interface, synthetic data generated by Gen AI is highlighted in light purple," he said. "We're very careful about how we display AI insights."

That caution reflects Nuix's deep roots in forensic investigation and regulation-heavy industries. Rubinsztein said the company's AI tools are particularly well suited to sectors where data governance is paramount - banking, insurance, pharmaceuticals, legal and law enforcement.

"When you're working with regulators and defence organisations, you have to be able to show your work," he said. "You have to prove how you got to a conclusion."

Looking ahead, Nuix is focusing on automation as the next key advantage, following the acquisition of an enterprise automation layer that lets clients run repeatable data processes.

"We're solving data curation problems at scale, repeatably, with speed," he said. "That's what sets us apart."

Ultimately, Rubinsztein sees Nuix not as an AI company per se, but as the essential infrastructure behind it.

"We're the picks and shovels of the AI rush," he stressed.