Most data science companies provide their clients with two sets of data. The first comes from descriptive analysis – describing data with charts and graphs based on what’s happened over the last year or two – and the second comes from predictive analytics, which predicts what will happen in the future using machine learning and Artificial Intelligence (AI).

Sydney-based retail predictive analytics and insights start-up, HIVERY is taking data science to the next level by adding an additional layer with prescriptive analysis – what should happen – removing human effort to understand the data, as well as human bias that sways future decision-making.

HIVERY chief technical officer, Andy McQuarrie and HIVERY co-founder and head of innovation, Matt Robards spoke at the recent Databricks ‘Driving Innovation with Data + AI’ APAC event on how they help companies generate a return on retail space investment.

Backed by The Coca-Cola Company, CSIRO, and Blackbird Ventures, HIVERY is pioneering hyper-local retailing. HIVERY uses data science and AI to offer hyperlocal retail recommendations on product assortment, space allocation and promotion.

“The core of everything we do is removing bias from the decision-making process by not only offering predictions, like other data science companies, but prescribing the best action you can take based on those predictions,” Robards said.

“Every company tries to generate a return on investment (ROI) but we specifically try to help companies generate a return on retail space investment – a term that we coined because companies, whether it be the smallest convenience store or the biggest retailer in the world, they invest a lot of money in space and that space comes at a real premium.

“With such a large investment, we need to ensure they have the right products in the space at the right price and with the right promotional activity behind it. Underpinning all of that is human-centred design.

“To achieve this, we have been building a suite of products because there is a broad range of different types of retail point of sale right from vending up to enormous hyper retailers. We have HIVERY Enhance which is our flagship product and first ever solution built around the vending space to achieve the right space and product allocation, up to HIVERY Curate is helping massive stores such as Walmart to figure out which products to put in the space and how much of that product to achieve the best efficiencies. HIVERY Promote is a tool to help optimise promotional calendars throughout the year.”

“Space equals money is an obvious thing to say but it’s a huge problem for a range of businesses. We believe optimising shelf space is a US$984 billion a year opportunity but it’s not a problem that can be easily solved using current approaches and the effort that prescriptive analysis goes to is not enough,” McQuarrie said.

“The HIVERY Platform Architecture is made up of Space, Promotion and Vending, using Microsoft cloud computing service, Azure and Azure Databricks that holds the machine learning and data science operations. It is more cost effective, faster and collaboration is easier.”