The Software Behind the Awkward Conversation About Money

Nobody enjoys a late-payment notice. The customer feels watched. The lender sees risk. The agent in the middle gets a screen full of dates, promises and rules, then has to make a decent decision before the next call arrives. EXUS builds the machinery behind that moment. Its software determines which account needs attention, what treatment is permitted, which channel has a chance of working and what happens if ordinary collections fail.

Founded in 1989 by George Konstantinidis, the London-headquartered company predates the web browser and the phrase “fintech.” It has survived by staying narrow. EXUS Financial Suite, or EFS, covers the debt lifecycle from early warning and delinquency through payment arrangements, field visits, legal recovery, write-off and debt sale. More than 9,500 people use it across over 50 countries, according to the company. Banks are the obvious buyers, but vehicle-finance companies, utilities and other large creditors have the same headache: many accounts, many rules, finite staff and customers whose circumstances do not fit cleanly into boxes.

Collections is a branching decision tree wearing a headset

The old caricature is a room of telephones. The actual operation is closer to air-traffic control. A missed payment may require a gentle reminder, an affordability conversation, a temporary arrangement, a specialist review, assignment to an outside agency or a legal process shaped by the jurisdiction and the collateral. Every action generates another piece of evidence. Every promise needs tracking. Regulators care not only that money is recovered, but how the customer was treated along the way.

EFS puts those branches in one system. The Collections module automates routine tasks and records calls, letters and digital messages. Legal Recoveries manages configurable court and asset processes, documents, fees, third parties and critical dates. Decision Intelligence segments portfolios and scores behavior. Digital Self Service gives customers a private, always-open route to submit information and choose approved repayment options. Field Collections sends mobile workers an optimised visit plan and keeps working when the signal disappears.

The useful distinction is not that EXUS automates. Every serious vendor does. It is that business users can draw and revise strategies with visual tools instead of commissioning a fresh development project. A lender can combine customer and account traits, set alternative treatments, test a new contact sequence and change course as behavior or regulation shifts. The software is meant to turn policy into something executable.

A five-year implementation met a one-week strategy cycle

Siam Commercial Bank offers the cleanest explanation of why this matters. When Colin Dinn became its chief technology officer in 2016, the bank was nearing the end of a five-year implementation of another collections platform. Yet agents still lacked a complete customer view. The project had delivered software without delivering the operational freedom Dinn wanted.

EXUS was not the safe-name choice. People inside the bank pushed for a larger vendor. Dinn changed his mind after reading an analyst report, adding EXUS to the tender, then flying to Athens to meet the team. What persuaded him was control. He said many suppliers maximised customisation in ways that made the client permanently dependent on them. EXUS demonstrated how the bank could tailor the system itself and gave a direct answer about delivery capacity.

“I don’t need a big name, I need the capability.”Colin Dinn, then CTO of Siam Commercial Bank

The reported result was specific: collections-team productivity improved 10 percent, while the time required to deliver a new strategy fell from a month to a week. The bank did not simply buy a more polished queue of accounts. It shortened the distance between deciding something and making the operation do it.

The same pattern appeared when National Bank of Bahrain brought Bahrain Islamic Bank into its group. NBB had an internal system with limited automation and tracking; BisB had an older EFS version. The mismatch showed up first as manual work, inconsistent records and thin reporting. The banks moved to a common, current version of EFS. Calls, letters and digital communications landed in one record, portfolio reporting became visible and BisB’s 13 collection strategies were simplified to three.

These stories reveal the failure that usually comes first: not the model, but the plumbing. Data sits in separate places. Teams cannot see the same history. Strategy becomes a meeting rather than an executable rule. When a merger or a new loan book arrives, the cracks become visible. EXUS’s product promise is to repair that operating layer before applying clever prediction to it.

The machine may suggest the next move. Policy still gets a vote.

EXUS now talks about machine learning, conversational assistants and agentic AI, but its more interesting claim is governance. Its decision tools can predict propensity to pay, likely contact time and channel, then propose a next action. In the company’s 2026 blueprint, the recommendation is checked against consent, recent contact, vulnerability, hardship and escalation rules. The result carries reasons and a timestamp. Models are monitored for drift, with evaluation gates and rollback plans.

That restraint matters. Collections is a terrible place for a black box with a lively imagination. A model can rank likely outcomes; it cannot decide that a regulator’s contact limit is optional or invent a repayment offer outside the lender’s policy. The practical design is a constrained loop: decisioning recommends, approved channels execute, interactions create structured signals and the next decision gets a little better informed.

Humans remain in the loop for disputes, vulnerability, high exposure and complex negotiation. EXUS is developing tools to summarise a case before a call, suggest compliant language during it and turn the transcript into coaching afterward. The ambition is not to remove the specialist so much as remove the scavenger hunt that occupies the first minutes of every difficult conversation.

One platform, local law, and no public price tag

EXUS sells enterprise software, implementation, support and specialist services. There is no public price list. The company promotes predictable costs and no upfront onboarding fee, but a buyer still has to scope modules, integrations, data migration, portfolios, jurisdictions and service levels. “No onboarding fee” is a procurement hook, not a promise that a multinational bank transformation is cheap.

The firm sits between specialist collections vendors and broad banking platforms. FICO, Pegasystems, Temenos, Qualco, C&R Software and in-house systems can all appear in the alternative column, though they do not make identical products. EXUS’s defence is domain density: one suite spans ordinary collections, external agencies, legal workflows, collateral, field visits, self-service and analytics. Global patterns are packaged into the product, while local partners and configurable processes handle market-specific rules.

That is also why it travels. Secure Trust Bank selected EFS for UK vehicle finance. First Bank has used it since 2006 and expanded from individual delinquency into more complicated micro and small-business portfolios. Bayport Finance used it to replace fragmented spreadsheets across six African countries. Pancreta Bank extended its relationship as acquired Greek banking portfolios were added. DLL runs the cloud version on Azure. The common problem is not a particular type of loan. It is operational complexity arriving faster than the old system can absorb it.

Start with the decision, not the demo

The transferable lesson is pleasingly untechnical. Map the decisions before shopping for features. Which signals trigger early intervention? Who can approve an arrangement? When must contact stop? What evidence needs to survive a complaint or court process? Which changes should an operations manager be able to make without IT? EXUS wins its best case studies when the customer knows the freedom it wants, not merely the screens it dislikes.

The conditions matter. A small creditor with one simple portfolio may be buying a cargo ship to cross a pond. A large one can still fail if customer data is unreliable, nobody owns the operating rules or local compliance is bolted on at the end. Predictive scores do not rescue a broken contact history. A configurable system does not help a team that cannot agree what to configure. And self-service does not work merely because a portal exists; customers need credible options, plain language and a channel they trust enough to use.

There is a quietly humane possibility inside all this machinery. Better segmentation can identify the customer who needs space, the one who forgot, the one who prefers a digital arrangement and the case that genuinely requires a person. Fewer pointless calls can lower cost and irritation at the same time. That outcome is not automatic, and it depends on the creditor’s policy as much as the vendor’s code. But it explains why EXUS calls collection workers “heroes.” The software handles the branching map. The person still has to navigate the conversation.