Saffron Building Society is using artificial intelligence (AI) to automate mortgage processing and increase “decision certainty” for its brokers.
The 180-year-old Essex-based mutual society has penned a deal with FintechOS to use the supplier’s AI platform and is getting started with its residential mortgage processing business.
The company plans to double its mortgage lending to £2bn over the next five years and so needed to increase its processing capacity through removing manual processes through AI.
Qasir Aslam, chief transformation officer at the company, described the residential mortgage project as being the company’s “first drop” into the IT platform as his team seeks to prove the value of AI on a case-by-case basis.
Changing expectations
The mutual society, which has approximately 200 staff with around 35 in the transformation team, is attempting to meet changing customer and broker expectations. The company sells all its mortgages via brokers, who require a clear understanding of the company’s offerings, while customers require a smooth process.
“From a broker perspective, they want to understand what our products are and who the customers are – then they need decision certainty,” Aslam told Computer Weekly. “Once information is submitted to us, how quickly can we get back to them with a decision principle and how quickly can we go from the offer to completion?”
Saffron deals with mortgages that are considered complicated. This includes customers who may not have an employment history, or their financial information might not be as straightforward.
“This could be self-employed people, it could be individuals or professionals who they don’t have their regular income, or it could be landlords, for example,” explained Aslam.
“When we implement our new platform, we want to make sure that we’re asking the right questions up-front at application capture, so that when we come back with the decision in principle, brokers have the confidence that it will go through.”
Staff at Saffron will interface with an AI agent known as DAX, with plans to also make this available to brokers. Aslam added: “Colleagues can ask DAX questions, but the plan going forward is then to make that available to brokers as well. If brokers want to understand something specific around a product and a customer, they could ask that agent.”
Aslam described Saffron’s plans to use the platform on a case-by-case basis: “We are going to create a continuous improvement work stream. As part of that, we’re looking at all the processing that we’ve got now, which isn’t optimised, and then looking at what tools are available to help us improve it. That’s where we’ll then look at whether AI will become one of those applicable tools for us.”
Drop by drop
This “drop-by-drop” approach, according to Aslam, will help his team prove the value of AI in use cases to the business: “The biggest challenge, based on my experience of implementing lots of these platforms, is making sure that we can implement something quickly that shows to the wider business the value that the platform can deliver.
“What I am trying to do with the business areas is [uncover] what the first drop of value that we can go and deploy early next year [is], so that we can demonstrate delivery success and show how great this platform is. We then want to iteratively build on it. Once we’ve deployed our first release, our first drop of value, we can add more products onto the platform.
“With residential mortgage origination, we want to lay the foundations so that the end-to-end application journey, the mortgage lifecycle, is on FintechOS. Once it’s live, we’ll learn a lot more about the platform and how we use the platform, then optimise and make it even more efficient,” added Aslam.
He said that while a company of Saffron’s size cannot match the processing efficiencies of its large competitors, “where we identify a process is broken, we’ll be able to vastly improve those”.
Aslam said the company had to ensure its staff came on the AI journey with the company: “The first thing we needed to do was reassure them. We’ve been clear from a Saffron perspective that when we look at AI, we’re going to make sure it’s human centric.
“The way I explain it to everybody is that AI isn’t there to take your job away. AI is there to help you to automate the mundane, day-to-day tasks so that you can focus on more value add.”
He said value add is about having time to go and speak to brokers, understand what their points are, and understand how to improve products. “That’s where we want to move the workforce and where they spend a lot of their brain power,” he said.
The use of AI is being emulated and encouraged across the mortgage sector.
Earlier this year, Vernon Building Society described its use of FintechOS, stating that it can now compete with big mortgage providers on the speed of mortgage processing, while ensuring its people focus on how it differentiates.
The Stockport-based company, which has 25,000 members and total assets of around £534m, specialises in savings and more complex mortgages. The firm is growing, with a 4.6% increase in mortgages on its books, now worth £439.5m, a 5.4% increase in total assets from last year.
Separately, in December 2025, as part of its plan to build the “mortgage market of the future”, the Financial Conduct Authority (FCA) outlined four areas of focus.
One of these, which looks at innovation and disclosure, will increase the use of AI among brokers while preserving humans in the process, according to the FCA. It said it is “encouraging the use of data and technology, such as AI, to help brokers give better and faster advice while keeping a human touch”.
