Primark, one of the UK’s major high street retailers, needed an urgent review of its entire property portfolio to establish liabilities on a site-by-site basis. With 189 stores to review nationwide in two weeks, the timescales were incredibly tight.

To add to the challenge, the leases varied hugely in scope, age, detail and exposure to risk. It was a complex puzzle where some stores were the sole responsibility of Primark, while others were managed by the landlord. We had to pull all the pieces together – and the clock was ticking.

We rapidly built an AI-system that could review each contract in minutes, not days. But to really get the most from the technology we needed to know which questions to ask. For Primark, this came down to two things: What is the definition of the premises we are responsible for? What does the repairing clause say?

We collated and fed the information into the system, asked the right questions, and within seconds we had the answers. The process also identified leases where issues were more complex, in which case we sent them to experienced real estate solicitors to assess and advise on the nuances and obligations. The results were then presented digitally, using pie charts, graphics and Excel sheets, to make it easy to share and understand.

Using a combination of tech and legal expertise, we were able to complete the review in days, rather than weeks, saving Primark 70% in time and costs. In the process, we have also created a valuable repository of information, bringing all the company’s disparate contracts together. It means that in the future we could easily find answers to other questions – whether it’s sub-letting restrictions or residential flexibility.

Perhaps most important of all, this project shows that AI is not just for multi-million pound projects of staggering scope, scale and cost. With the right knowledge, AI can rapidly be applied to everyday tasks, delivering real-world savings very rapidly.