EXCLUSIVE: "Mortgages Minus the Monolith" - Adam Archibald, Mambu, Ben Ussher-Stanley, nCino and Sumitpal Ghuman, Publicis Sapient in ' The Fintech Magazine'

Technology that does the ordinary while freeing people to do the extraordinary was what Mambu, nCino and Publicis Sapient came together to provide for specialist UK lender OSB Group and its hundreds of intermediaries

The UK mortgage market is characterised by complexity. From its uniquely short refinancing lifecycle to having one of the most diverse borrower profiles of any major economy. Landlords run portfolios of houses in multiple occupation (HMOs) and multi-unit blocks through limited companies. Self-employed borrowers juggle patchwork incomes and less-than-perfect credit history.

Older borrowers want terms that stretch well past retirement. Large loans, second charge loans, bridging loans, shared ownership, Right to Buy... and nearly all of the applications for these arrive through a broker whose entire job is finding the one lender with an appetite for their client’s unique set of circumstances.

“In a mortgage ecosystem like the UK’s, there are literally millions of home borrowers out there who don’t fit into a nice neat little box,” says Ben Ussher-Stanley, Principal Solutions Specialist at lending software provider nCino. Which is why lenders who might once have been thought ‘specialist’ are now considered mainstream.
One industry prediction is that this type of bespoke lending will have increased by 70 per cent between 2023 and 2029. Research from Landmark Information Group found that 85 per cent of lenders now regard differentiation as essential to success, up from 59 per cent in 2024, with more of them tailoring propositions to specific customer segments.

And that requires a sophisticated blend of  the digital and the human.

“A borrower needs a person who knows what they’re doing, who can sit and look at their case and analyse and understand it,” adds Ussher-Stanley.
The same research found that the length of transactions has overtaken regulation as lenders’ biggest frustration, an issue cited by 40 per cent of those surveyed. This points to a genuinely thorny problem. The more bespoke the loan, the more complex the risk assessment and underwriting becomes. Yet brokers – who place the overwhelming majority of UK mortgage business – want speed and certainty for every client, however unusual their circumstances. 

Meanwhile, the market metrics against which lending happens never stay still. UK Finance expects around 1.8 million fixed-rate deals to mature this year against a backdrop of stretched affordability, and forecasts that new buy-to-let lending will stay flat as landlords absorb heavier taxes and regulation.

A lender built for complexity 

Few lenders are more aware of thesetensions than OSB Group. A FTSE 250 specialist lender, with a net loan book of £25.9billion at the end of 2025, it works exclusively through intermediaries while dealing directly with savers through subsidiary brands including Kent Reliance and Charter Savings Bank.

It was the UK’s largest independent buy-to-let lender by gross new lending in 2024, and the lending territory OSB inhabits is precisely the kind that makes high street banks break out in a cold sweat. In 2022 it embarked on a five-year programme to rebuild its technology, appointing Publicis Sapient as its exclusive end-to-end transformation partner. Publicis Sapient, in turn, helped OSB assemble the specialists around it: Mambu, whose Cloud-native, API-first core handles products and servicing and makes the underlying data available to everything layered on top, and nCino, whose platform runs origination, workflow and the broker-facing experience.

The first milestone, a digital savings platform built on Mambu’s microservices core, went live in October 2024. The lending platform followed in 2025, and OSB used the moment to tidy up a brand structure that had grown tangled through acquisitions. Buy-to-let lending, previously split between Kent Reliance for Intermediaries and Precise, was consolidated into a new brand, Rely, which launched officially in November last year after a pilot involving more than 50 broker firms. 

Precise now concentrates on residential and bridging loans; InterBay focusses on commercial lending; Rely targets everyone from first-time landlords to large portfolio investors, explicitly courting limited companies, HMOs and multi-unit blocks. Presenting full-year results in March, OSB Chief Executive Andy Golding said the programme’s milestones had been hit on schedule and within budget. At Rely’s launch, Group Intermediary Director Adrian Moloney summed up the thinking behind the platform as using technology ‘to do the ordinary, whilst freeing our people to do the extraordinary’.

That thinking starts with recognising who the customer is – which is not, as you might assume, the person actually buying the house.

“Realistically, in this market the customer outcome is determined by the brokers,” says Ussher-Stanley. The nCino experience is therefore built around the broker, albeit working with the borrower, to complete the application and submit documents, without having to phone the lender for guidance. Where data already exists elsewhere, it is absorbed and used to pre-populate forms. The aim is to reach a decision as quickly as possible, because certainty – be it a fast ‘yes’ or a fast ‘no’ – is what the borrower really needs.

Sumitpal Ghuman, UK Mortgages Product Lead at Publicis Sapient, says the programme started not with technology but with business strategy, which in practice meant mapping what the broker was trying to achieve.

“The broker is not trying to understand what the platform is, nor what the architecture is behind it. The broker is looking at speed of application, confidence, clarity, and the outcomes that they need from the products they’re looking for, for their customers.

“Transformation should be measured not in good architecture diagrams, but in the impact it is having on the broker.”

On OSB’s platform, he says, brokers can now register in under four minutes and get an agreement in principle (AIP) ‘with certainty in under 10 minutes’, while the journey from AIP to offer takes ‘meaningfully less time’ than before.

Plumbing before paint

Ussher-Stanley started his career in banking and saw how money was poured into front ends after the neobanks came to market with their fancy apps.

“The customer said ‘this is amazing – it’s modern, sleek, elegant’. But in the background there was an absolute mess of spaghetti [systems].”
nCino reverses that logic, he says: “The plumbing is the most important part of this.”

In mortgages, that means everything the underwriter relies on arrives as clean data, which means they can concentrate on the nuance in deciding whether to green-light a loan and under what terms. That’s distinct from a computer-says-no scenario, and it plays to the human ability to bring real-world experience to a case.

“With many underwriters, their talent is the instinct they have about a case,” says Ghuman. “But in a world where they spend three hours keying data in rather than three hours analysing it, they are not able to use that wisdom all that often.”

For Adam Archibald, Regional Director for the UK and Ireland at Mambu, the plumbing problem is partly due to how banks used to buy technology: one monolithic system from one vendor, crammed with as much functionality as possible, which reduced risk and meant they didn’t have to ‘worry about things like integration’. Cloud, APIs and AI broke that model.

“We’re now seeing organisations wanting to decouple services. They want to know more about microservices architecture and how to use middleware components,” he says. “But while that all makes sense on paper, if your software partners don’t have modern APIs, then you’re going to struggle. The beauty of this partnership with nCino and Publicis Sapient is that our solutions are designed on modern APIs.”

What lenders like OSB get by abandoning the monolith is agility and lower running costs, Archibald says. Where a simple rate change on a legacy stack could take weeks or months, with SaaS and no-code configuration, some Mambu clients now make product changes in days. And where product and pricing rules are duplicated across front-end and back-end systems, a single source of truth removes the manual reconciliation that adds risk and cost.

Just as important in a world where technology is constantly updating, those lenders still on legacy solutions ‘will find they miss the opportunities that others with more modern technologies are able to capture’, says Archibald.

As a back-office solution, Mambu isn’t really involved in the customer experience.

“Organisations who use our technology are interested in it being resilient and reliable, with the ability to access data at speed,” adds Archibald. “The reason the partnership works so well with nCino and Publicis Sapient is that Mambu provides access to quality data at pace, so they can deliver great customer experience.”

OSB’s upgrade spanned buy-to-let, residential and commercial mortgages, and involved Cloud migration and new middleware, but it was purposefully delivered incrementally. First, the partners shaped the architecture, then proved it worked before scaling. In a UK market, where so much depends on a lender’s credibility with brokers, that mattered.

They had to be comfortable using it before it was universally adopted. That foundation is also what makes the next technology phase in lending possible. Landmark’s research also found that three-quarters of lenders expect AI to enhance customer engagement and 68 per cent expect it to improve data analysis and decisioning. But, as Archibald says, that optimism is ‘predicated on having quality data and having good access to that data’.

With agentic workflows gaining ground in credit underwriting, Mambu has implemented support for the Model Context Protocol (MCP) – the open standard, originated by Anthropic in late 2024, that lets AI agents connect automatically to software rather than needing a bespoke integration for every pairing. But the ultimate aim, Archibald says, is to pair AI processing with human processing so – as OSB’s chief said – people are free to do what they do best. Over time, Mambu hopes to automate more journeys with partners such as nCino and Publicis Sapient.

Ussher-Stanley sees nCino’s role shifting accordingly. Three years ago, he says, automation meant API call-outs and ‘if-this-then-that’ workflows. Now nCino is becoming an orchestration layer, the ‘layer through which your agent might be able to talk to my agent’. Crucially for a regulated lender like OSB, it is also the governance layer, where how it came to a decision is clearly identifiable. Fill that layer with internal and external data, overlay AI, says Ussher-Stanley, and ‘suddenly you have a completely different ecosystem in terms of how you’re doing lending’.

Making the switch

The benefits to the lender and the borrower don’t end at completion. A mortgage in the UK runs for 25 or 30 years, and keeping good payers on the book at every renewal cycle is firmly in the lender’s interest, especially as interest rates lower, optimism rises and the market becomes more bullish. With servicing data from Mambu feeding into nCino, a lender can not only spot early signs that a borrower might be heading into financial stress, but also that they might have capacity for a further advance.

“When your rate comes up for renewal, why do we put so much stress onto people? Because actually, from the technology standpoint, that should be pretty straightforward,” says Ussher-Stanley, who describes a scenario where a borrower is contacted six months ahead of the renewal date with a set of options. 
“Within three clicks of a button they can probably have their mortgage,” he says.

UK Finance estimates that internal product transfers, where borrowers switch to a new deal with their existing lender rather than remortgage elsewhere, will have grown two per cent (to £261billion) in 2026, with the vast majority of borrowers staying put. But it also forecast a 10 per cent increase in external remortgaging activity, driven by intensifying competition among lenders and recent rule changes that simplify switching.

The human touch

None of the technology solutions that Mambu, nCino and Publicis Sapient promote are about hollowing out the workforce, insists Ussher-Stanley. In fact, it’s precisely the opposite in an industry where specialism is all to do with human judgement. Too often, he says, skilled staff spend their time cutting and pasting data ‘from point A to point B, and then point B to point C’. What distinguishes one lender from another, in his view, is not only the processes it chooses to orchestrate, its policy and risk appetite, but ‘most critically, who are your people?’.

Nobody builds it all

The fact that three technology partners were involved in the OSB project illustrates how an ecosystem approach secures best-of-breed components for the lender. No single vendor can excel at everything. nCino believes it has the leading mortgage origination solution in the UK, but readily acknowledges that it requires partner technology around it, which ‘we have no aspiration or intention to build, because it’s not within our wheelhouse’.

The pace of change in both mortgage lending and lending technology demands this more specialised approach in both sectors. Asked on a FinTechTalk panel what lending will look like in five years, Ussher-Stanley’s honest answer was that nobody can say. 

“But whatever that future is, nCino is architected in such a way that if something new comes up and you say, ‘I want a piece of that’, we’re going to be able to say, ‘you can absolutely have it’.”

Ghuman defines future-ready lending in similar terms. A lender should not need to ‘reinvent the entire architecture if there is a change in market, a change in regulation, or if the distribution changes’.

That means adopting Cloud, API-first and composable architecture, and a strong data layer. Publicis Sapient’s job then is to convert whatever Mambu and nCino release next into value for the lender before the competition gets there.

“What we’ve found with our partnership with nCino and Publicis Sapient is that by delivering technology in a decoupled way you start to unlock the true value of APIs, Cloud and AI,” says Mambu’s Archibald. “Here in the UK, but also in other regions, mortgage providers and lenders are seeing the value of having best-of-breed solutions in specific areas. So, if your secret sauce is credit underwriting, how can you find a solution that makes sense for that? If you’re looking to have better data access for things like AI, how do you find a solution that fits that model? The only way is to have a more decoupled approach.”

For the borrower who doesn’t fit a neat box, the result should be a broker who gets a straight answer quickly, an underwriter with time to think, and a lender that makes contact well in advance of the fixed rate ending with an offer that’s built around you.