Tarocash owner Retail Apparel Group to give HR systems an AI boost

Retail Apparel Group, owner of menswear stores like Tarocash, Conner and .yd, is preparing to serve up AI within its human resources function amid a company-wide push to take advantage of the technology.

Head of people and pay Nicole O’Dowd Martins said that using AI was the next step for the company after having spent five years consolidating disparate payroll and rostering system used across its six retail brands into one.

O’Dowd Martins told attendees at the Dayforce Summit that the company was planning to start by making its human resources systems more self-serve for its workforce of just under 5000 employees that keep its 600 stores open.

Prior to moving onto Dayforce’s human resources platform, the clothing brands operator had been using very primitive ERP systems and even relied on WhatsApp chat groups to notify staff of their rostered hours.

Dayforce has delivered a much-needed modernisation but it also created some additional cognitive load for staff.

It's allowed the company to create a large library of human resources documentation and training-related content. However, staff have difficulty finding what they need and teaching them how to locate it has become a burden, according to O’Dowd Martins.

“We do spend a lot of time instructing and directing the team on how to access what they need or answering questions that are very basic, so enabling our team to self-serve a lot of that will absolutely free our team up and a lot of senior leaders to focus on what they need to,” she said.

RAG already makes use of some common AI tools such as ChatGPT and Microsoft Copilot-style AI to reap productivity gains on day-to-day tasks, O’Dowd Martins said.

However, the AI to be deployed within the people and division’s environment will rely on native AI capability Dayforce has developed specifically for people management.

O’Dowd Martins said that it was expected to include both agentic components and chatbot-style query systems.

“Wherever we have team that have access to certain data and [appropriate] security clearances within our system, they'll be able to access anything that sits within there,” she said.

It’s not clear precisely which flavour of AI technology that will provide the underlying smarts for the system.

However, Dayforce says that it uses multiple models to build its AI capability, including a mix of open-source large language models, managed Databricks-hosted AI infrastructure, and proprietary machine learning models for analytics and workflows.

O’Dowd Martins said that the business case for deploying the technology in RAG’s people and pay system environment was only one of many that its dedicated AI team was currently assessing and prioritising.

“We're now exploring what are the opportunities right across the business as to where to activate AI solutions within systems that we're already using or to introduce them as new components,” she said.