Woolworths lifts the covers on internal AI
Key points
- Woolworths has revealed it uses an AI tool called Team Assist to field 7000 employee inquiries per week, escalating only one in 10 to a human.
- An AI marketing tool has cut production of the weekly shopper catalogue from a week to a few hours, chief executive Amanda Bardwell said.
- Woolworths’ head of SAP Murtuza Imdadali disclosed two more AI tools being trialled, a training chatbot for its promotions app and a system for measuring inventory dimensions.
Woolworths has started to disclose its inward-focused efforts to use AI, in the last month alone revealing how it uses the technology in two areas of the business it hasn’t previously discussed publicly.
Chief executive Amanda Bardwell took the opportunity of the group’s full year results announcement to reveal that supermarket giant is using two AI tools for employee assistance and to accelerate its weekly marketing campaigns.
“We believe AI will help deliver better experiences for customers, for teams, and will help us transform our operations and workflows,” Bardwell said.
Both are having a “real impact”, according to Bardwell.
An AI employee support tool called ‘Team Assist’ acts as a decision-making aid that is designed to remove the need for using humans to take employee inquiries.
Bardwell said that Team Assist fielded 7000 employee inquires per week in the last financial year, with the need for escalation to a human removed in nine out of 10 cases.
Woolworths’ newly-revealed AI marketing tool appears not to have been given a friendly name like Team Assist, but the company said it had sped up production of its weekly shopper catalogue “from a week to a few hours”.
Two more AI tools break cover
In addition to these two tools, Woolworths’ head of SAP Murtuza Imdadali revealed that the supermarket giant is quietly trialling at least two more AI tools at a software industry event held earlier this month.
One of these, a training chatbot, also appears destined for the hands of Woolworths’ marketing staff.
Imdadali said the training chatbot, which guides staff on how to use its promotions app, is “hopefully going live soon”.
Like Team Assist, the ultimate goal of the training chatbot, he said, is to reduce load on its support operations.
Imdadali said that Woolworths’ promotions app is “complex” and required “a learning curve” to use it effectively.
Instead of calling support or approaching colleagues, staff could use the chatbot to surface answers from years of training materials that Woolworths has stored away in its corporate data.
“A user who gets stuck on a particular screen or needs some understanding of how to fill in some of the questions that are being asked or what are the expectations, they can get an immediate response. Presently, when they get stuck, they generally log a service call. It goes to our support teams, or they ask their colleagues, I assume.
“The good side effect of that is the user experience turnaround time is quicker. They get instant response, but also reduces the support calls to the support teams in if they were stuck in a scenario,” Imdadali said.
The other tool Imdadali is working on automates the process of measuring the dimensions of inventory, appears more likely to be used in its supply chain operations.
It is for spotting inaccuracies in the dimensions of the thousands of products that retailer sells.
He said that the inaccuracies can impact “many parts” of Woolworths’ operations such ranging, selling and deliveries.
“We’re still trying to tune it for the AI to identify if a particular, I don’t know a Mars Bar or Snickers … have the correct dimensions.
“We’re doing in it a way to give a sense of confidence: low, medium, high.
“[If it’s] high that it’s incorrect, then that gets flagged, and we’re working out whether we automate that process or whether we set it as a report or workflow, then for the user to then go through and accept the new values or review it and update the values,” Imdadali explained.
Until now, most of the focus on Woolworths AI capability has focused on its customer-facing efforts, such as its Olive agentic shopping assistant and its “smart basket” – the latter being able to predict shopper needs and automatically add items to baskets.
Broader digital results
The company said engagement across its digital channels is strengthening, with weekly visits to its smart phone application and website increasing 23 percent compared to last financial year.
It said that ecommerce orders using its app exceeded 70 percent for the first time in the fourth quarter of the 2026 financial year with the ecommerce group’s overall sales increasing 16.9 percent.
Net profit for the financial year was up 18.1 percent to $1.14 billion; it would have been higher but for a program of remediating historical underpayments to staff.
About Author
