The company has built an AI that helps companies discover which insights are important to a particular business question— no matter how vast or complex the datasets analyzed. Large consumer companies use Unsupervised to better understand their customers, so they can optimize key parts of their business.
With numerous successful Fortune 500 deployments, Unsupervised isn’t just a technical innovation that discovers new insights—it’s also the easiest way for companies to get real value from AI. Customer testimonials highlight companies using Unsupervised to successfully integrate AI into their analysis and decision-making process in weeks, not years.
Their next-gen AI scours complex datasets to unearth patterns and metrics, without human supervision. “Unsupervised automatically discovers which insights matter from the variety of data types your company already has. This allows companies to better understand their customers, and to use this understanding to improve their key business challenges,” asserts Tyler Willis, Co-founder and Chief Product Officer of Unsupervised.
Discussing how companies act on insights, Willis underscores a crucial point in the debate around AI and automation: “It’s widely-accepted that AI can process a lot more information than the human brain, but it cannot replace human understanding or decision-making. AI can’t infer context about your business, so it can’t make good decisions for you. Our AI serves you important insights, so you can see what’s important and decide how to act on it.”
Unsupervised uses unsupervised learning to analyze a data-set and discover the relevant and nuanced patterns that drive customer behavior. Customers define a business problem they want to better understand, and Unsupervised finds the important, related patterns in the data. Their AI can analyze all types of data—structured, unstructured, incomplete, first or third party—and it has been used by companies across many industries, including CPG, retail, telecom, and financial services.
Behind Unsupervised’s compelling value proposition is the unique experience of its co-founders, Tyler Willis and Noah Horton. Both are industry veterans with deep knowledge in advanced AI and consumer marketing. The duo met while building Involver, which launched the social media management software category. Co-founded by Horton, Involver grew to power a large number of Fortune 1000 companies before it was acquired by Oracle and turned into the nucleus of Oracle Marketing Cloud. Horton went on to lead Oracle’s Public Cloud efforts. While Horton was climbing the ranks at Oracle, Willis became a consumer CMO, leading marketing teams towards becoming more customer-centric and data-driven.
“At the onset, we realized that the more complex your data is, the harder it is to find what matters,” explains Willis. With experience as the head of business and corporate development for a top data company, and later as a consumer CMO, he’d seen the problem first hand. “Discovering patterns in high-dimensional, messy data lets us deliver real value to business users and data scientists, both of whom work better with fewer dimensions.
If we find what matters, they can start the process already having many of the answers they need.”
Unsupervised automatically discovers which insights matter from the variety of data types your company already has
The company’s distinct approach to data is symbolic of the next level in AI. Seeing high-dimensionality data as an asset, instead of a problem, dramatically changes how companies think about finding insights.
To better understand how it works, we asked Willis to walk us through a use case, he shared this example: “a large retailer uses Unsupervised to analyze their product and transaction data. They find complex groups of products with the biggest increases or decreases in sales. Their merchandising team looks at these trends and decides which products to promote and which to discount.” The types of patterns that Unsupervised found indicates the power in being able to use complex data, “you see obvious groups, like ‘menswear,’ but you also see incredibly nuanced groups like ‘Grey sweaters described as ruffled’—and because the AI sorts patterns by importance, you only see groups that can meaningfully impact revenue.”
An example of building with customers in mind, Unsupervised easily integrates with an organization’s existing architecture. Users can access all insights through a user-friendly dashboard as well as export output for additional validation or for use in automated systems. In addition, the Unsupervised system is built around robust security and data protection standards. For maximum client benefit, the company also offers services around data augmentation, technical integrations, and analysis. “Long story short, our company is designed around a ‘never-let-a-customer-fail model,’” quips Willis.
Unsupervised brings a unique and marked shift in AI. In just a year, the company has successfully showcased their value in the enterprise. When asked about their roadmap, Willis refused to comment on upcoming customer announcements and product improvements but foreshadowed some: “We’re doing something fundamentally new. When you can look at data complexity as an asset, it completely changes the types of solutions you build for data analysis.”