Albert Heijn revolutionizes the store employee experience using Azure OpenAI

“We all have a digital standard in our day-to-day lives which sets our expectation of being able to do seamlessly on a mobile device. That is what employees expect of us. So we constantly asked ourselves – how do we bring that standard into our store employees’ day-to-day work to help them with daily tasks?”

Jelle Gijsbers, Product Manager at Albert Heijn, is explaining the driving force behind the supermarket chain’s latest initiative. As a frontrunner in AI adoption, the company uses innovative technology for customer-facing initiatives and more recently, for employee initiatives too.   

“In the retail industry,” continues Gijsbers, “we still have the labor-intensive process of stocking shelves and at Albert Heijn, this job is usually completed by minors who work a couple of hours a week after school. As they're relatively young and working fewer hours, we set ourselves the goal of making our store employees’ work easier.”

Albert Heijn previously launched an employee app, @AH, designed to enhance its digital experience for its employees. With the help of Microsoft technology, the supermarket has taken a next step to better reflect the needs of its workforce while also positively impacting satisfaction among the workforce in their day-to-day jobs.

Paving a way for a more tech savvy workforce 

Albert Heijn is one of the largest supermarket chains in the Netherlands. Its presence is so big in the country, that around one in four Dutch people are estimated to have worked there at some point in their lives or work there currently, and many of them gain their first experience of work at the retail chain. For its 700 owned stores across the Netherlands, it employs over 80,000 people who work on the shop floor, interacting with the millions of customers that come through its doors each day. 

“The broader challenge that many businesses are facing today, including us, is that the labor market is getting tighter,” explains Gijsbers. “The age of our store employees is getting younger over time and talent retention is crucial. In the long run, this isn’t sustainable. So, we need to ensure that we are optimizing the work we’re doing.”

To do this, Albert Heijn focused on finding ways to make the store employee experience more relevant and inclusive for all ages. “I myself started at the company when I was 16,” begins Gijsbers, “and having that personal experience of working in the store I understand and empathize with the struggles they face.”

“These 14–16-year-olds come after school for a couple of hours to help stock the stores,” continues Gijsbers, “but often they’re approached by customers with questions or they’re unsure where a product needs to be restocked. It’s a hurdle for them to ask management for the answers to queries because they don’t want to come across as untrained or unaware. It’s all a bit daunting for them.”

Gijsbers and his team saw this problem as an opportunity to transform the store employee experience and see how generative AI could support its entire store workforce.

A conversational assistant is envisioned

Albert Heijn is no stranger to innovation. Last year it became the first Dutch supermarket to use generative AI to make the customer experience more convenient with its AH customer App, a smart tool that provides recipes and tips for cooking and storing dishes.

“We already had a foundation built using Microsoft technology,” says Gijsbers, “so we set ourselves a goal: to find a way to leverage the enablement platform to make the labor-intensive processes our store workforce encounter more efficient by 1%.”

The supermarket chain teamed up with Microsoft once again and its partner EPAM to come up with a strategy. “We started with a traditional product approach where we all came together for ideation workshops to explore the challenges from a business and employee perspective,” says Gijsbers.

“We created personas, used empathy mapping and looked at the frictions people at the company experience,” adds Gijsbers. “We defined a few hypotheses and then looked to validate them. Based on our findings, Microsoft suggested we build a virtual assistant that could answer questions that store employees are confronted with while on shift.”

Using low-fi prototyping, the team imitated a conversational AI solution with store management to see how it could work. “In the backroom of the store,” explains Gijsbers, “we had five people sitting around computers with our website open while holding handheld devices. Employees ask a question, and we quickly type back the answer. We did it in three stores for one shift and it was really explorative, helping us to focus on truly valuable use cases for the end users.”

“As we were developing the concept,” says Gijsbers, “we tried to get feedback from those who would be using the conversational assistant to check that each part of the solution was helpful. For example, we were originally going to make the tool speech-to-text or speech-to-speech. But our store employees said they found talking to their phones awkward and even a bit rude in front of customers – so we listened and rescoped the experience.”

Having validated the concept of a virtual assistant, it paved the way to build the minimum viable product. 

“This collaboration with Albert Heijn was about more than just technology; it was enabling a cultural shift by building an AI assistant hand-in-hand with employees, ensuring it met the needs and inspired adoption by a Gen Z workforce,” explains Raman Bakanovich, Account Director at EPAM. 

“One of the success factors was establishing a robust technical foundation with the AH Gen AI enablement platform. This not only makes implementation of AI-enabled business solutions faster, cheaper, and safer but also provides Albert Heijn with an efficient base for future AI initiatives,” he adds.

Building on a proven foundation of technology

The conversational assistant was developed on top of the central generative AI platform, similar to the supermarket’s customer conversational assistant. It was hosted on Azure Kubernetes Service, a solution that could deploy and scale the app. All the company’s data was stored using Azure Database PostgreSQL, a fully managed database that can integrate with AI tools at a low cost. 

“Some of the capabilities that we built up in that foundational layer of technology we were able to reuse,” says Gijsbers. “We found we could optimize the original model, reducing the cost to build.”

The assistant lives on the employee-facing @AH app, which is accessed on the personal mobile devices of the staff. “Everybody understands a conversational way of interacting with software, so we didn’t need to train anyone to use the assistant, and using personal devices added that familiarity.” says Gijsbers.

Using Azure OpenAI, the conversational assistant is able to answer a whole range of questions from what needs to be restocked on shelves to where a product can be found in store. 

“Barcodes are a really important means of identifying a product,” says Gijsbers, “so we also added in a barcode scanner. There’s also a store floorplan and a planogram of the shelves.”

“Assistants have mainly been used to optimize office work or to handle return shipments, for example,” adds Gijsbers, “but our sole focus has been on supporting store employees in their day-to-day tasks by addressing their specific needs.”

Sparking a cultural evolution

So far, the assistant has been piloted in eight stores, with a view to scale up to 25 stores in summer 2025 and has been met with a lot of positive feedback with store employees saying the technology is easy to use. Its success is, in part, down to the store employee's willingness to adopt the new technology and their openness to a cultural transformation. 

“Adoption is definitely a huge challenge when it comes to a new solution,” says Gijsbers. “There has been a lot of discussion around whether it’s okay for employees to be walking around the store looking at their phones. But employees are on their phones all the time outside of work and we need to adapt to these generational nuances.”

“The majority of people are behind it and customers don’t see an issue with it, especially since store employees are able to answer their queries much faster now,” continues Gijsbers. “Introducing this assistant proposition has helped us to become an attractive employer and we’re able to retain staff because they’re excited to work for us now that we have the solutions in place that makes their work experience better.”

“It will take time for store employees to get used to this new way of working, and to encourage them to use the AI assistant as a first line response,” adds Gijsbers. “But the great thing about technology like this, is that it's always adapting and as we find new use cases, the app will become more intuitive and an integral part of store employees’ day-to-day working lives.”

Looking ahead to the assistant’s untapped potential 

The conversational assistant is now cemented in the architecture of the company and the supermarket chain hopes to roll out the technology to all its 700 owned stores later this year. Since piloting the conversational assistant, Albert Heijn has noticed an increase in productivity among its store employees and while it’s too early to see the impact on job satisfaction and talent retention, it looks promising.

“We’re continually looking at ways to improve the solution and address the internal issues our employees face,” says Gijsbers. “We’re always finding new use cases and we want to move some of our focus to helping our store management teams and also adding multiple languages to the app.”

“We also want to explore the app’s metanalysis capabilities,” says Gijsbers. “There’s an infinite number of unexplored questions we could get feedback on, and we need to be able to process it and understand where we need to make improvements. This technology offers us so much potential to innovate.”

“From the beginning,” concludes Gijsbers, “the eagerness from the Microsoft team and EPAM to try and make a success of this initiative and their collaboration has made this possible. It’s been amazing to be able to bring our concept to life and improve the work of our all our store employees to ultimately deliver on our mission: ‘Together we make eating better the easy choice, for everyone’.”