Allica: using tech to put relationships back at the heart of SME banking

For Allica Bank, winning the Excellence in Technology Award, sponsored by VIP Apps Consulting, at the Finance Connect Summer Awards 2026 brought recognition of a technology strategy that has been built firmly around the needs of brokers and customers.

But for Niv Subramanian, Deputy CEO of Allica Bank, awards are only one measure of whether that strategy is succeeding.

“Winning any award, particularly from an industry body that is quite relevant to our strategy, is always nice,” she said. “Given how hard the team works, it is nice recognition to say that we’re moving in the right direction.”

However, she stressed that the most important measure remains the experience of those using Allica’s products.

“The feedback that we care about most is what we hear from our brokers and our customers. Making their lives easier and meeting their needs better is what drives our product development.”

Automating behind the relationship

That customer-led approach is central to Allica’s philosophy towards technology.

Rather than viewing digitalisation and relationship banking as competing models, Subramanian believes the two need to work together – particularly when serving established SMEs.

Established SMEs – those firms with 5-250 employees that Allica is built specifically to serve – can be too complex for a completely standardised digital journey, but the economics of the market also make it difficult to deliver an entirely manual, people-intensive service at scale. The answer, she argues, is to use technology to support the relationship rather than replace it.

“Our philosophy is that you automate behind the relationship,” she explained. “The best technology should be where it’s almost invisible.”

For the customer or broker, the result should not necessarily be an awareness that AI is operating behind the scenes. Instead, they should experience a bank that knows more about them, responds faster and gives its relationship teams more time to engage with them.

This distinction is particularly important as the financial services industry debates whether greater automation will inevitably make banking less personal.

Subramanian believes the opposite could be true.

Investment in automation can make relationship banking more economically viable by reducing the cost and administrative burden behind the scenes while retaining people at the front end.

“If you invest properly in technology automation behind the scenes, it allows us to have people in the front end,” she said. “Perversely, it is not taking away the humanness. It is actually bringing it back.”

What does AI mean for brokers?

For all the attention surrounding generative AI and emerging technologies, Subramanian believes the immediate opportunity for asset finance brokers is relatively straightforward: removing friction.

Rather than focusing on highly visible AI applications or “magical automation”, she sees some of the greatest value coming from reducing the effort involved in assembling information, processing documents, checking applications and chasing lenders.

“For brokers, ultimately, the AI that matters most is not some glamorous chatbot that does everything,” she said. “It is removing friction, taking out their own effort so they can put that effort towards customers.”

In an industry where speed is crucial, this can translate into faster and more consistent responses without requiring brokers to do additional work.

Allica is already seeing this in practice.

The bank has developed a lending agent capable of reading unstructured information such as an email from a broker, assessing whether the required information has been provided and, where appropriate, passing the case through to decisioning.

In some cases, that means Allica can respond to a broker in around ten minutes, compared with a process that might previously have required somebody to pick up and review the case before it could progress – a process that may previously have taken some banks days or weeks.

But speed is not the only benefit.

Subramanian highlighted consistency as another important consideration for brokers. Greater automation can create a more consistent underlying approach, with experienced underwriters then applying judgement where it adds most value.

Technology is also helping Allica broaden its proposition. By reducing the administrative work required of underwriters, the bank can devote more resources to complex cases and expand its appetite into areas that would previously have been more time-consuming to assess.

The best indication that this is working, according to Subramanian, is not brokers telling Allica they like its technology.

“The feedback from them is not, ‘we love your tech’,” she said. “It is, ‘it’s quicker and easier’. That’s the feedback and that’s what we value.”

AI supporting – not replacing – judgement

While AI can increasingly deal with unstructured information and help with more complex cases, Subramanian is clear that Allica is not seeking to remove human judgement from lending.

Generative AI is not currently used by the bank to make lending decisions. Instead, it can help interpret information and assemble a case for an underwriter, while more traditional LLM models can automate straightforward decisions.

“We still don’t use Gen AI for decisioning, but we use it to help with putting together a package of what it looks like for the underwriter,” she explained.

This allows experienced underwriters to spend less time on straightforward cases and more time on the “gnarly”, bespoke transactions.

Complex company structures, unusual circumstances and cases requiring an assessment of mitigating factors are not always reducible to rules. For Subramanian, that means experienced professionals will continue to have an important role, even as AI takes on considerably more of the heavy lifting.

The same principle applies elsewhere in the lending journey. AI agents can undertake elements of operational processes, including supplier checks, while a person retains oversight and the final say.

Removing repetitive administration also has implications for the quality of roles within the industry. Rather than spending time transferring information between documents and systems, employees can concentrate increasingly on analysis, critical thinking and decision-making.

Giving relationship managers more time with customers

Allica is applying the same philosophy to its relationship teams.

AI tools can help relationship managers research customers and prepare for meetings in a fraction of the time previously required, producing information and discussion points that enable them to have more informed conversations.

The objective is straightforward: less time preparing at a desk and more time with customers and brokers.

“It has definitely allowed us more time to actually be with brokers,” Subramanian said.

This also explains why Allica has resisted using technology simply to push customers or brokers towards automated channels.

The bank does not, for example, use a customer-facing AI chatbot. Customers can use Allica’s app or online banking if that is their preference, while brokers may choose to communicate by email. But the important principle is that the individual retains a choice over how they interact with the bank.

Allica’s technology is therefore intended to make every channel work more effectively rather than forcing customers away from human interaction.

Responsible AI and the importance of trust

As AI becomes more embedded within financial services, governance, transparency and explainability will become increasingly important.

Subramanian said Allica sets high accuracy thresholds when introducing AI tools and continues to monitor models once they are in use, recognising that their performance can change over time.

Crucially, Allica also asks AI systems to provide the rationale behind their outputs.

“It is exceptionally important to make sure that this is fair and transparent and explainable,” she said.

She also cautioned against holding AI to an unrealistic standard of perfection.

Human decision-makers make mistakes too, she pointed out, and the appropriate comparison should therefore be between the performance and risks of AI and those of the existing process – rather than assuming technology can only be trusted if it is infallible.

There are significant issues around areas such as security that require careful management, but Subramanian believes expectations around AI need to reflect the reality that neither humans nor technology will achieve 100% accuracy.

The next opportunity: ‘always-on’ credit

Looking further ahead, one of the most interesting opportunities Subramanian identified is the potential development of an “always-on” credit view of customers.

That could be particularly relevant in asset finance, where businesses regularly replace assets and return for funding.

Rather than relying primarily on credit assessments refreshed periodically, AI could potentially draw together financial data with behavioural information, news and other signals to maintain a much more dynamic picture of a customer.

If successfully developed, Subramanian believes that could ultimately allow an existing customer to access funding within minutes.

AI could also provide lenders with a more real-time view across their portfolios, identifying emerging patterns in demand and risk.

It is part of a wider future in which much of the machinery surrounding lending could increasingly disappear into the background.

There will remain practical hurdles – including access to open data and customers’ willingness to allow lenders to connect directly to their information – but the direction of travel is towards reducing the administrative work surrounding finance and allowing brokers to devote more of their time to customers.

Making relationship banking possible at scale

In corporate banking, large organisations can afford sector specialists covering areas such as agriculture, manufacturing and retail. That model has traditionally been much harder to replicate economically for established SMEs.

AI has the potential to change that equation.

Technology can give relationship managers access to relevant sector knowledge and insight within minutes, enabling them to have more informed and tailored conversations without needing to be dedicated specialists themselves.

“Technology just allows you to make relationship banking possible at scale,” said Subramanian. “You can have a much more meaningful customer conversation. You can be a lot more tailored to your customer.”

For established SMEs – and ultimately the wider economy – she believes that could be hugely valuable.

And that brings Allica’s technology strategy back to where it started.

Success is not ultimately measured by how much AI the bank deploys, or how visible its technology becomes. Instead, Subramanian wants technology to enable higher customer and broker satisfaction, continued growth and better-quality lending decisions.

For brokers, the results should be tangible: simpler applications, quicker decisions, more consistent outcomes, less effort and more time with their BDMs.

“Tech enables the business,” she concluded. “Brokers and customers should expect to see continued improvement in how they engage with us.

“They should see us improving their lives and making it simpler.”