Al Karim Somji and the price of changing a bank

A bank can remember your balance to the penny and still struggle to recognise you across its own products. You may have a mortgage, a savings account and a credit card, yet each can inhabit a separate administrative universe. The institution knows a great deal about your money. Getting its systems to agree on what your custom deserves is another matter.

Al Karim Somji built a company around that awkward distinction. Zafin, which he co-founded in 2002, gave banks a way to manage products and pricing outside the older systems that processed their business. Behind the technical language sat a familiar commercial question: if a customer brings more business, can the bank offer a better deal?

His career has followed that question deeper into the machinery. He led Zafin until April 2024, founded another banking technology business, OpenCoreOS, and joined FIS after it acquired the startup in March 2026. The companies and titles have changed. The practical difficulty has proved unusually durable: banks need to improve systems that customers expect to work throughout the improvements.

Before the bank, a shop counter

Somji was born and raised in Tanzania. Before university, he helped run his family’s retail business and worked on a restaurant, a bookstore and a stationery shop. These were businesses in which a customer relationship had a visible human scale: somebody came through the door, bought something and might come back.

He later moved to Canada and studied computer engineering at the University of Ottawa. His early Canadian jobs included selling ice cream and washing dishes while waiting to begin school. The route into enterprise software had a decidedly practical introduction. There were customers long before there were bank technology committees.

His early professional career included Bell Canada, with work spanning business development, sales and corporate engineering. In 1997 he founded Eftia OSS Solutions, developing operational support systems for telecommunications companies. That put him inside the industry whose ambitions would supply the seed of Zafin.

The combination matters. Somji had experience with both the engineering beneath a service and the commercial work of selling it. Later, when he talked about banking technology, the conversation kept returning to what an institution could actually offer a customer. A system’s usefulness had to survive contact with a transaction.

The wrong industry, the right problem

Zafin began with an interest in telecom. The arrival of 3G suggested that telephone companies could distribute new services, including financial products. Somji studied the way Japan’s NTT DoCoMo earned money from paid content passing through its network and considered what a similar model would require elsewhere.

Following the payments led him toward banks. There he encountered products and prices embedded in older systems, with customer information divided among them. A bank might want to reward the whole relationship, but the software had been organised around individual products. A commercial idea could become an expensive engineering assignment.

By 2013 he was explaining relationship pricing as a means of recognising the business a customer already gave a bank. Transparency belonged in the proposition: customers should understand the fees they paid and the benefits they received. Loyalty becomes a rather eccentric request when the institution cannot explain the bargain.

The insight travelled from one industry to another, then acquired a different purpose. Telecom had provided the initial opportunity. Banking supplied a persistent obstacle worth building around. Zafin’s eventual identity came from paying attention to that obstacle rather than remaining attached to the market where the founders started.

The patient business of paying for software

The founders initially funded development through other work and a professional services business. Services supplied revenue and direct experience of customer requirements. Before Zafin could package a repeatable product, it had to learn what banks would ask that product to do.

The change from bespoke work to a product required money as well as engineering. Somji used debt financing while the business developed the structure needed for outside equity. Kayne Anderson invested in 2014. The company began its cloud transition in 2016, addressing the complexity of supporting different installations and versions.

It is easy to tell a founder’s story as a succession of fundraising announcements. Somji’s account gives more attention to the work between them: separating product development from delivery, establishing governance and choosing when to sell ownership. Those decisions are less photogenic than a cheque and considerably harder to reverse.

His advice in 2022 was blunt: “Don’t look for money when you urgently need it.” A funding plan, in his view, should anticipate the cost of development, recruitment and new markets. Cash ought to arrive before the next stage becomes a matter of urgency. Even ambition benefits from making an appointment.

“Don’t look for money when you urgently need it.”Al Karim Somji / 2022

An alliance, an acquisition, a change of mind

In December 2018, Accenture announced an alliance with Zafin and an agreement to acquire selected professional services assets. Some Zafin employees would move to Accenture, which would become a preferred integrator of the software. Somji described the arrangement as a way to concentrate on product innovation and simplify Zafin’s operating model.

The division of work made the commercial logic tangible. Zafin could develop its platform while a much larger services organisation helped banks put it to use. Enterprise software has an unglamorous final requirement: somebody must make it work inside an institution with its own history, habits and systems.

In January 2022, Zafin bought FinancialCAD, known as FINCAD, for US$32.7 million. The acquisition brought quantitative analytics into the picture. Somji was interested in applying capabilities associated with capital markets to banking pricing. The combined business was reported at the time to have 500 employees.

FINCAD’s stay was brief. In April 2023, Numerix announced that it had acquired the business from Zafin. Somji said Zafin had worked on FINCAD’s product and customer success strategies and that Numerix offered it a route for further growth. The episode is a useful reminder that a company’s course includes both purchases and departures.

Twenty-two years, then a different chair

In 2023, Zafin won Microsoft’s Financial Services Global Partner of the Year award. The recognition concerned customer solutions built on Microsoft technology, a concrete milestone for a business that had spent years moving toward cloud delivery. Its work had become part of a larger banking technology ecosystem.

In February 2024, Nordic Capital announced an agreement to acquire a majority share in Zafin. The founders and management would reinvest alongside it. The customer list included Wells Fargo, HSBC, ING, CIBC, PNC and ANZ: institutions operating in different markets, with the shared difficulty of changing established technology.

The deal’s rationale followed the original product idea. Separating products and pricing from multiple core systems could help banks introduce offers while maintaining operational control. The buyer was investing in a business built around a carefully chosen boundary inside a bank.

That April, Charbel Safadi succeeded Somji as CEO. Somji remained a board member and substantial owner. In his farewell message as chief executive, he thanked co-founders Anugopal Venugopalan and Dinesh Krishnan, the team, customers and his family. He expressed confidence in Safadi’s leadership. A 22-year tenure ended with the founder keeping a seat at the table and passing over the daily job.

Back to the foundations

OpenCoreOS brought Somji back to company building in 2025. Its founding team included Steve Van Wyk, formerly HSBC’s global chief information officer, and technology leaders Slavo Vojacek and Ricky Marcon. The new venture took aim at the core banking platform itself, extending his work beyond the product and pricing layer.

The company presented an AI-native platform designed to operate across multiple clouds. Its proposition included real-time ledgering and tools for creating financial products. Those were statements of what the team was building and offering; the essential promise was continuity while a bank changed the technology underneath its business.

Somji’s BIAN board role connects that ambition with standards. BIAN, the Banking Industry Architecture Network, works on shared banking architecture. His stated interest there is in standards-based, open-source platforms that can support core modernisation. A bank’s systems need ways to cooperate if individual components are to change.

There is a consistent commercial instinct in this work. Somji keeps returning to constraints that are expensive for an individual institution to remove. A shared platform can turn one bank’s difficult internal project into something other banks can also use. The challenge is persuading those institutions to trust the machinery with their everyday business.

The bank stays open

FIS acquired OpenCoreOS in March 2026. Somji subsequently became its president of enterprise platforms. The acquisition brought his latest venture inside an established banking technology provider, moving the next phase of his work into a company already supplying the systems banks use to operate.

In September, FIS described an enterprise platform strategy intended to let institutions modernise progressively while preserving core stability. That is a corporate objective with a familiar shape to anyone following Somji’s career: improve what a bank can do without demanding that it stop being a bank while the work takes place.

Asked about his guiding principles in 2022, Somji named “Integrity and resiliency”. His explanation linked both to relationships with colleagues, partners and customers, and to carrying on when business became difficult. The words fit a field where engineering promises have to last through long implementations and institutions have good reasons to be cautious.

The distance from a family shop in Tanzania to bank infrastructure at FIS is considerable. Yet a recognisable question runs through the journey. What does a business know about its customer, and can it act on that knowledge? Somji has spent decades working through the software consequences. The customer, reasonably enough, would just like the account to work and the deal to make sense.