Peak looks to improve retail AI project success rates

Peak announced its AI Platform-as-a-Service (PaaS) Beta programme at AWS re:Invent in Las Vegas this week.

Richard Potter, CEO of Peak, said at the event: “Research shows that AI projects often fail due to the complexity of deployment within retail. Every retailer must become AI-driven, yet no system is dedicated to this in retail today. What’s more, we know 80% of a data team’s time is spent cleaning or collecting data, data manipulation, building training sets or with feature extraction. Our AI system changes that, and with it data teams are able to focus on what they do best.”

Legacy business systems are inflexible, slow and expensive to maintain, he continued. Data warehouses are not built to deliver machine learning solutions and data tends to exist in silos due to the proliferation of cloud applications. This all makes AI deployment challenging for many businesses.

He claimed that the venture’s AI System is delivering powerful results for brands such as Asos and FootAsylum, in some cases improving revenues from marketing by 28%. “We’ve launched our PaaS Beta programme to enable enterprises with existing data science teams to use the same system. With it they will improve their machine learning capability to extract more value from retail data and experience the real benefits of AI.”