Fixing the £460 Million Monster: How We’re Using AI to Augment, Not Replace, Human Expertise in Mortgages

SInce entering the UK mortgage industry, I’ve realised it’s fundamentally archaic. For most people, getting a mortgage is the most stressful financial transaction of their lives. It’s a time sink where people spend dozens of hours lost and confused; a “black box” where unhappiness stems from simply not knowing what’s happening; and a “document nightmare” that leads to delays, collapsed property chains, and crushing disappointment. 

This isn’t a niche problem. It’s a monster that feeds on inertia and fear, costing UK consumers a staggering £460 million a year in mortgage overpayments. That’s thousands of pounds stolen from a family’s budget, simply because the process of switching is too painful to face. 

As leaders in this space, we face a trifecta of challenges: intense margin pressure, rising customer expectations, and a heavy compliance burden. We have a responsibility to drag this process squarely into the 21st century. The motivation to solve this is immense, and it’s the mission that drives my team. 

The AI tsunami is upon us, and while it presents a huge opportunity for service businesses like mortgage brokerage firms, it also brings a wave of hype. I want to share our practical, on-the-ground playbook for using AI not as a silver bullet, but as a powerful tool to augment our most valuable asset: our human mortgage experts. 

First, We Had to Measure the Pain 

When I first arrived at our brokerage, I realised the team had done a phenomenal job of scaling the business, but from an operational perspective, we were flying blind. We thought we knew where the time was going, but the data told a different story. 

This became our foundational principle: you can’t fix what you can’t measure. 

Our first step wasn’t to buy a suite of AI tools; it was to build the muscle to understand our own business. We invested in Process Mining and Capacity Management to get a true, data-driven picture of every step in our operation. We started asking the hard questions: 

• How long does it really take for a case manager to check a lender’s offer for accuracy? 

• How many minutes are spent chasing solicitors for an update? 

• Where are the hidden bottlenecks that frustrate our team and delay our customers? 

Only by meticulously measuring the pain could we identify the right problems to solve. This data-first approach gave us our roadmap and ensured that our investment in technology would be targeted at areas of genuine, measurable impact. This is the first step in any battle for “AI-ready data” – understanding your own operational reality. 

Our Guiding Principle: Augment, Don’t Replace 

The cost of intelligence is heading to zero. Tasks we once believed only humans could do—like reading and interpreting complex documents—AI can now do. The cost of building software is collapsing, and where coding was once the exclusive domain of engineers, today, English is becoming a programming language. 

This changes everything. 

AI allows us to build internal tools that non-engineers can use – tools that take expertise, embed it, and automate it to create huge productivity gains. And that means freeing our teams to do the things that actually matter: 

• Building trust with customers. 

• Guiding people through high-stakes decisions. 

• Handling complex cases with real care. 

The firms that will win in this new world are the ones with deep experience and deep data. Experience becomes training data. Training data becomes capability. And capability becomes a competitive advantage. 

That’s why our principle is simple: Augment, don’t replace. 

Every tool we build is a piece of an “Ironman Suit” for our advisors – something that makes them faster, smarter, more capable, and, crucially, more human in the moments that count. This is the future. Not dystopian robo-advice, but a new generation of super-powered, bionic advisors. That’s how we believe brokers win and retain customers for life. 

The Playbook in Action: A Tale of Six Innovations 

Our strategy is focused on two fronts: freeing up our experts to be experts, and creating a friction-free customer experience. Here are six practical examples of how we’ve put our philosophy into action. 

Part 1: Freeing Up Our Experts 

We’re not making our advisors work harder; we’re removing the tasks that get in the way of them giving great advice. 

• Case Study 1: Automated QA. The traditional method of quality assurance is a highly manual, inefficient process. Selecting 5-10% of calls at random to check for compliance creates a culture of fear and is prone to misinterpretation. Our solution was to use AI to analyse 100% of customer calls against our compliance framework, stitching together multiple interactions to see the full picture. The outcome? We moved from random spot-checks to holistic coaching, cut our QA admin time by 90%, and improved the quality of the customer experience. 

• Case Study 2: Smart Booking. Our most valuable asset is our advisors’ time. Allocating a standard 60-minute slot for a very simple remortgage is wasted capacity. We developed triage technology that identifies “low-complexity” cases upfront and offers shorter, 30-minute booking slots. The outcome was a 20% increase in our total daily appointment capacity, allowing our advisors to help more customers. 

• Case Study 3: Automated Form Filling. Not all lenders offer modern API connections, meaning our team had to juggle dozens of different portals to manually key in applications—a process taking, on average, 45 minutes. We used a Large Language Model (LLM) to automate the population of these portals directly from our core CRM. This single initiative has cut our average time to key applications by over half. 

Part 2: Creating a Friction-Free Customer Experience 

• Case Study 4: Nurturing New Leads. Not every lead is ready for a full advice call. Pushing them too early leads to drop-off. We built an automated nurture sequence that provides valuable content—guides, checklists, calculators—to top-of-funnel customers, helping them progress at their own pace. This simple change led to a 20% increase in our lead-to-submission conversion rate. 

• Case Study 5: Automated Offer Checking. A case manager’s daily task is to manually confirm the accuracy of a lender’s offer letter against our internal records. This took around 10 minutes per case. We employed AI to triangulate the data between the lender’s offer, the official ESIS document, and our CRM, with a human-in-the-loop to handle discrepancies. Today, over 98% of this process is fully automated, catching errors earlier and reducing customer stress. 

• Case Study 6: The Rate Checker. This is our proudest addition. Once a customer locks in a rate, completion can take weeks. During this time, rates can drop. Manually tracking this is impossible at scale. We built an automated rate checker that constantly scans for lower rates with the same lender during the application process. If a better deal appears, it alerts the customer and, with their consent, switches them. The outcome? We’re saving our customers over £6 million a year. This builds incredible trust and turns a moment of potential anxiety into a moment of delight. 

Our Three Lessons for Smart Growth 

This journey has been one of practical, iterative learning. To draw to a close, I’ll leave you with the three lessons we’ve learned that underpin building a better future in our industry. 

1. Lay the Foundations: You can’t fix what you can’t measure. Before you rush into AI, invest in the tools and the discipline to truly understand your own operational data. Find the pain points, quantify them, and let the data be your guide. 
2. Give Your Experts an Ironman Suit: The true productivity paradox of GenAI is solved when you focus on augmentation, not replacement. Focus your technology on removing the 80% of administrative burden so your human experts can be freed up to do the high-value, trust-building work that only they can do. 
3. Obsess Over the Customer: Don’t just automate for efficiency’s sake. Be smart about how you use technology to create moments of trust and delight. An innovation like our Rate Checker doesn’t just save money; it proves to the customer that you are unequivocally on their side, long after the initial sale is made. 

That is how I believe service businesses will win in the age of AI.