EP1007: Award-winning core banking

This interview features Myles Bertrand discussing how Tuum is helping modernize core banking systems through flexible and cloud-based solutions. The discussion highlights Tuum’s successful collaboration with LHV, which shifted the bank’s IT focus from costly maintenance toward innovation-driven growth. Bertrand explains the importance of the company’s cloud-agnostic strategy and its partnership with Google Cloud in supporting scalable and future-ready banking operations. The interview also explores Tuum’s expansion into Islamic banking and SME financial services, emphasizing the need for adaptable digital infrastructure. Additionally, Bertrand advocates for progressive modernization, where banks upgrade systems gradually instead of replacing them entirely. Overall, the interview shows how API-first architectures and AI-ready platforms help banks remain competitive in an evolving financial landscape.

Transcript

SPEAKER_01 0:00

Right now, you know, your smartphone can use artificial intelligence to instantly map a route across the entire globe. Or um recommend this highly personalized playlist and even translate six languages simultaneously.

SPEAKER_00 0:12

Oh yeah. The tech we carry in our pockets every day is just staggering.

SPEAKER_01 0:16

It really is. But then, and this is what we're getting into today, why does it still take your traditional bank like three full business days to clear a simple paper check?

SPEAKER_00 0:28

Yeah, it's incredibly frustrating.

SPEAKER_01 0:29

Right. Or when they finally roll out a new mobile feature, why does it look like a website built in, I don't know, 1998? We interact with these flawless, sleek interfaces all day long, yet the moment we touch our life savings, it feels like we're stepping back in time.

SPEAKER_00 0:44

The contrast is definitely jarring. And um, it actually stems from a massive hidden infrastructure problem. The industry refers to this as core banking.

SPEAKER_01 0:52

Cru banking.

SPEAKER_00 0:53

Yeah, it is the foundational digital vault, you know, the central ledger. Basically, the absolute bottom layer plumbing of every single financial transaction you make.

SPEAKER_01 1:02

And upgrading that plumbing is the exact mission for today's deep dive. We are exploring a really fascinating March 2025 interview from the IBSI FinTech Journal.

SPEAKER_00 1:13

Right. The sit-down with Miles Batran.

SPEAKER_01 1:15

Aaron Powell Exactly. He's the CEO of a banking technology provider called TUMA. And um this interview really sheds light on how massive, highly regulated banks are attempting what is essentially a heart and lung transplant on their legacy systems.

SPEAKER_00 1:29

Which is not an easy operation.

SPEAKER_01 1:31

Not at all. They are trying to stop merely surviving their own outdated tech and you know, start actually innovating.

SPEAKER_00 1:37

And the source material provides this brilliant case study of what happens when a bank actually pulls this off. Because Tumum collaborated with a client called LHV, and they jointly won Best Core Banking Installation at the Global FinTech Innovation Awards.

SPEAKER_01 1:50

Wow.

SPEAKER_00 1:50

Yeah. The award is prestigious, sure, but the underlying data point is what really matters here. Bitran revealed that before this whole transformation, LHV was spending 80% of its massive IT budget solely on maintenance.

SPEAKER_01 2:04

I mean, I always picture legacy bank IT like a terrifying ceiling-high game of Jenga.

SPEAKER_00 2:10

That's a pretty good way to look at it.

SPEAKER_01 2:11

Right. Like every time you want to pull out a block, maybe uh you want to update the mobile app's login screen or something, you threaten to collapse the 50-year-old checking account block resting right underneath it.

SPEAKER_00 2:22

Exactly. Everything is tangled together.

SPEAKER_01 2:24

Aaron Powell So 80% of your budget is just hiring engineers to stand around holding the tower steady, totally terrified to breathe on it, which leaves what? Almost nothing to actually build new cool features.

SPEAKER_00 2:38

That visual is incredibly accurate. Because when your engineers are just consumed by patching holes and managing brittle code, they aren't building the future. But by implementing Toom's modern architecture, LHV managed a complete inversion.

SPEAKER_01 2:54

Wait, they flipped it.

SPEAKER_00 2:55

They flipped the ratio completely. So they now spend 80% of their IT budget on innovation and only 20% on maintenance.

SPEAKER_01 3:03

Okay. See, flipping to 80% innovation sounds almost like a fantasy for a bank. I'm trying to figure out the actual mechanics of that. Because if you are terrified of knocking over the Jenga Tower, you can't just bulldoze the tower overnight and drop a brand new Lego set on the table.

SPEAKER_00 3:17

Aaron Powell No, definitely not.

SPEAKER_01 3:19

Aaron Powell I mean, in banking, if you break the core ledger, people's mortgages don't get paid. The ATM shutdown, complete system crash is a catastrophic event.

SPEAKER_00 3:27

Aaron Powell Right. And that fear of a system crash is exactly why most bank executives paralyze their own organizations. The traditional method for upgrading these massive databases is known in the industry as rip and replace.

SPEAKER_01 3:41

Sounds violent.

SPEAKER_00 3:42

It kind of is. The bank builds a massive new system over several years, picks a weekend, turns the old mainframe off, turns the new one on, and just holds their breath.

SPEAKER_01 3:52

Oh man, that sounds incredibly stressful.

SPEAKER_00 3:54

It is a notorious high-risk strategy that, well, it frequently ends in highly public failures. So Tua VM advocates for a completely different methodology, which they call progressive modernization.

SPEAKER_01 4:04

Progressive modernization. Okay, let me guess how this works in practice. To avoid the weekend blackout, the bank just runs the old 1980s mainframe and the shiny new cloud software at the exact same time.

SPEAKER_00 4:17

Well, not exactly.

SPEAKER_01 4:18

But wait, if my checking account data is split across two completely different systems that don't even speak the same coding language, doesn't that reconciliation process cost double? You're basically paying to maintain the crumbling Jenga Tower while also building the Lego set.

SPEAKER_00 4:34

Yeah, running parallel duplicate systems for the exact same customer accounts would be a chaotic nightmare. That isn't how progressive modernization functions mechanically. The entire strategy is really built around risk isolation.

SPEAKER_01 4:47

Okay, risk isolation. How so?

SPEAKER_00 4:49

Instead of splitting existing data, the bank identifies one highly specific brand new priority. Like perhaps they want to launch a spin-off digital bank for a younger demographic, or maybe they need to enter a new European market or just roll out a specialized commercial lending product.

SPEAKER_01 5:05

Oh, I see. So they isolate the new initiative entirely.

SPEAKER_00 5:08

Exactly.

SPEAKER_01 5:08

The old legacy mainframe keeps managing my everyday checking count completely undisturbed. But the brand new digital lending app is built and launched exclusively on the new Tuim core.

SPEAKER_00 5:18

You've got it. Because building that new digital product on top of the old rigid core would have taken years and millions of dollars just to untangle the code.

SPEAKER_01 5:28

Right, you'd be messing with the Jenga blocks again.

SPEAKER_00 5:30

Precisely. By launching it on a fresh modern core, the bank delivers immediate value to the market. They start generating revenue from that new product in months, not years.

SPEAKER_01 5:40

That makes a lot of sense.

SPEAKER_00 5:42

And most crucially, the tech team builds internal momentum. They prove to a highly skeptical board of directors that the new technology is secure and more importantly, profitable.

SPEAKER_01 5:52

Yeah, you prove the concept in a safe, contained sandbox before you migrate the entire population.

SPEAKER_00 5:58

Aaron Powell Exactly, one manageable chunk at a time.

SPEAKER_01 6:00

Aaron Powell Okay, but to make that modular flexible approach work, I'm assuming you can't rely on the bank's old physical server room. You need the cloud.

SPEAKER_00 6:09

Absolutely. The cloud is mandatory for this.

SPEAKER_01 6:11

Aaron Powell And the interview mentions that Toom is cloud agnostic, meaning, you know, a bank could use Amazon Web Services or Microsoft Azure, but it specifically highlights a very crucial partnership with Google Cloud. Why zero in on Google?

SPEAKER_00 6:24

So the Google Cloud Partnership is a strategic imperative for certain global regions, particularly the Middle East. Cloud adoption is accelerating rapidly in that market, but banks operate under severe regulatory scrutiny regarding data sovereignty.

SPEAKER_01 6:39

Data sovereignty, meaning the physical servers holding the financial data must literally reside within that country's geographic borders.

SPEAKER_00 6:47

Exactly.

SPEAKER_01 6:48

Like you can't process a Saudi citizens' mortgage application on a server sitting in a warehouse in Ohio.

SPEAKER_00 6:53

That is the core issue, yes. Google Clarence has invested heavily in building physical infrastructure within these specific regions. By partnering with Google, Toom ensures that banks can utilize highly advanced cloud architecture while remaining strictly compliant with local data sovereignty laws.

SPEAKER_01 7:09

That is a huge advantage.

SPEAKER_00 7:11

It is. And getting onto the cloud isn't just about storage efficiency anymore. It is the absolute prerequisite for the next major mandate in financial services, artificial intelligence.

SPEAKER_01 7:21

Right. AI. This is where we hit the technical wall with legacy banking.

SPEAKER_00 7:24

Yeah.

SPEAKER_01 7:25

Because AI is fundamentally reshaping how we analyze data. But an AI model is really only as smart as the information it can consume.

SPEAKER_00 7:34

Right. It needs data.

SPEAKER_01 7:35

Yeah. For an AI to instantly approve a complex loan or give a customer real-time advice on their spending habits, it needs real-time, uninterrupted access to the bank's data.

SPEAKER_00 7:46

Aaron Powell And the reality check for traditional banks is that their legacy systems literally go offline. Many of these older cores still operate on a mechanism called batch processing.

SPEAKER_01 7:56

Aaron Powell Batch processing. Okay, let me make sure I understand this correctly. The bank system collects all the individual transactions, like ATM withdrawals, cleared checks throughout the entire day in a sort of digital holding pen.

SPEAKER_00 8:07

Yeah, that's a good way to describe it.

SPEAKER_01 8:09

Then at midnight, the system essentially shuts down to the outside world and spends the next five hours processing that massive batch of data, updating the central ledger, and reconciling all the accounts at once.

SPEAKER_00 8:21

That is precisely the mechanism. And during that five-hour window, the core ledger is essentially locked down.

SPEAKER_01 8:26

Okay, so if we think about AI as this genius, world-class chef that you hired to cook for you 24 hours a day, batch processing is like throwing a heavy steel padlock on the chef's refrigerator for five hours every single night.

SPEAKER_00 8:41

Oh, I love that analogy.

SPEAKER_01 8:42

Right. The chef is wide awake, incredibly capable, and ready to work, but they're entirely useless because they cannot access any of the ingredients.

SPEAKER_00 8:51

A brilliant way to visualize the limitation. You just cannot have an AI agent offering dynamic second-by-second financial insights to a user if the database is literally asleep. This is why the Souls emphasizes that Tuum utilizes an API-first cloud native architecture.

SPEAKER_01 9:08

Aaron Powell Okay, let's break those two terms down for the listener because they definitely get thrown around a lot in tech circles. API stands for Application Programming Interface.

SPEAKER_00 9:15

Yes.

SPEAKER_01 9:15

I like to think of an API as the standardized menu at a restaurant. It allows two completely different software programs to talk to each other safely. The AI agent doesn't need to know how the bank's database is structured. It just looks at the API menu and says, you know, please hand me the current account balance for user A, and the API retrieves it instantly.

SPEAKER_00 9:35

The API is essentially the universal translator. In an API first system, the core ledger just securely holds the truth. And any authorized external application, whether it's an AI model, a mobile app, or a partner fintech company, can query that truth instantly without having to untangle any legacy code. Right. And then the cloud native part means the system was built from day one to operate in the cloud utilizing microservices.

SPEAKER_01 10:01

Meaning if a million people check their bank accounts on payday, a cloud native system automatically spins up thousands of tiny temporary servers to handle that massive traffic spike and then just deletes them when the rush is over.

SPEAKER_00 10:13

Exactly.

SPEAKER_01 10:14

Whereas a traditional physical mainframe would just get overwhelmed and probably crash.

SPEAKER_00 10:17

Right. That elasticity is what creates an always-on environment. The padlock is permanently removed from the refrigerator, to use your analogy.

SPEAKER_01 10:26

Which is great news for the chef.

SPEAKER_00 10:28

Exactly. Now having an always awake database is clearly a massive win for a bank's software engineers, but the interview delves into how this architecture actually solves intractable real-world problems. The source highlights two major areas here: Islamic banking and SME banking.

SPEAKER_01 10:45

Aaron Powell Okay. Let's start with Islamic banking. Tuam launched specific solutions for this sector in the fourth quarter of 2024.

SPEAKER_00 10:53

They did, yeah.

SPEAKER_01 10:53

Now, my understanding of Islamic finance is that it operates under strict ethical and religious compliance rules, most notably the absolute prohibition of charging or paying standard interest. So how does upgrading the software architecture actually address religious compliance?

SPEAKER_00 11:09

The challenge really lies in the computation complexity of the ledger. A traditional Western banking core is hard-coded to execute a very simple mathematical function. It takes a principal amount, applies an interest rate, and compounds it over time.

SPEAKER_01 11:21

Pretty straightforward.

SPEAKER_00 11:22

Right. But Islamic finance operates on fundamentally different mechanisms, such as profit and loss sharing agreements or multi-party asset ownership.

SPEAKER_01 11:31

So instead of a bank just lending you cash to buy a house and charging you 5% interest, the bank might actually purchase the house outright and then sell it back to you at a markup through a structured payment plan.

SPEAKER_00 11:41

That is a common structure called murabaha. And to manage that legally and ethically, the bank's software must track the underlying physical asset, figure out complex profit sharing ratios, and maintain a strict separation of funds to ensure money doesn't mix with prohibited industries.

SPEAKER_01 11:57

Oh wow, that sounds complicated for a computer from the 80s.

SPEAKER_00 12:00

It is a rigid legacy database from 1985, literally lacks the data columns and relationship structures to track those complex contracts natively, so engineers end up building incredibly fragile workarounds. But Toom's modern flexible ledger can be configured to process these complex sharia compliant contracts computationally in real time.

SPEAKER_01 12:20

And the demand for this is massive. The interview notes that regions like the Gulf Cooperation Council and Saudi Arabia in particular are moving at incredible speeds.

SPEAKER_00 12:30

Unprecedented speeds, really.

SPEAKER_01 12:31

Yeah, they aren't just updating old tech, they are leapfrogging entire generations of legacy software. They are demanding the exact same tap-and-go, frictionless digital experience as any tech hub in the world while still maintaining absolute ethical compliance.

SPEAKER_00 12:48

Miles Bitran refers to this as bringing a fintech mindset to conservative banking sectors. It allows institutions to scale incredibly fast because the software natively understands the rules of their specific financial ecosystem.

SPEAKER_01 13:02

That covers a massive regional shift. But the interview also points to a structural economic problem that affects almost every listener SME banking. Small and medium enterprises.

SPEAKER_00 13:11

Yes, a huge global pain point.

SPEAKER_01 13:14

If you run a small business, or if you've ever tried to secure a loan for a local bakery or a specialized consulting firm, you know the process is an absolute nightmare of manual paperwork.

SPEAKER_00 13:24

Traditional banks have notoriously struggled to serve the SME market profitably. I mean, small businesses are the lifeblood of global economies, but their financial profiles are highly idiosyncratic.

SPEAKER_01 13:36

Okay, I understand the paperwork is annoying, but does a faster cloud native database actually make a struggling bakery less risky to lend to? Like if their cupcakes aren't selling, better banking software doesn't magically fix their revenue. Why does modern core tech solve the SME lending problem?

SPEAKER_00 13:55

That's a great question. The underlying risk of the bakery failing remains exactly the same. The technology does not change the risk, it radically changes the cost of assessing that risk.

SPEAKER_01 14:05

Ah, okay. The cost to figure out if they're risky.

SPEAKER_00 14:08

Exactly. Under the old model, assessing a small business is highly labor-intensive. A mom and pop bakery doesn't fit neatly into a standard consumer credit score algorithm, but they also aren't a massive corporation with audited financial statements.

SPEAKER_01 14:20

Right, they fall on this weird middle ground.

SPEAKER_00 14:22

Yes. So a human loan officer has to sit at a desk, manually gather two years of tax returns, dig through printed bank statements, cross-reference inventory logs.

SPEAKER_01 14:30

Which takes forever.

SPEAKER_00 14:31

It requires dozens of hours of expensive human labor. So if the bakery is only asking for a $10,000 line of credit, the bank might spend $2,000 in human hours just verifying the paperwork.

SPEAKER_01 14:44

Wow, so the profit margin is completely erased before the loan is even issued.

SPEAKER_00 14:48

Exactly. The cost to serve is simply too high. But with a modern core, the bank utilizes a concept called embedded finance.

SPEAKER_01 14:57

Embedded finance.

SPEAKER_00 14:58

Embedded finance changes the entire underwriting paradigm. Instead of the bakery owner gathering documents and bringing them to the bank, the bank's API plugs directly into the software the bakery already uses to run their business.

SPEAKER_01 15:11

Oh, that's brilliant.

SPEAKER_00 15:12

Right. The bank integrates seamlessly with the bakery's point of sale system, their digital accounting software, or their e-commerce dashboard.

SPEAKER_01 15:20

So the bank's AI can just analyze the bakery's daily cash flow, verify their revenue streams, and assess their seasonal dips in milliseconds.

SPEAKER_00 15:29

Exactly.

SPEAKER_01 15:29

And the cost of underwriting that $10,000 loan drops from $2,000 in human labor to basically a fraction of a penny in computing power.

SPEAKER_00 15:38

The profit margin returns. Suddenly, a traditional bank can serve millions of small businesses profitably at scale. They can instantly offer tailored products such as buy now, pay later options for B2B purchases, or instant multi-currency accounts for an online retailer sourcing materials globally? The technology completely transforms the unit economics of lending.

SPEAKER_01 15:59

So if the benefits are this transformative, you know, if you can flip your IT budget from 80% maintenance to innovation, if you can unlock always-on AI, and if you can suddenly lend to small businesses profitably at scale, why isn't every major bank in the world immediately migrating to a modern core like Tuam?

SPEAKER_00 16:17

Aaron Powell Well, the barrier is rarely the technology itself. The true hurdle is legacy thinking. The interview provides a very pointed warning regarding this mindset, highlighting the failure of HSBC's Zing project.

SPEAKER_01 16:29

Oh, the Zing project. That is the perfect cautionary tale. HSBC is a massive established global bank, and they recognize the threat from those nimble international payment apps, the fast-moving fintechs that let you move money across borders cheaply. Right. So HSBC decided to build their own competitor, an app called Zing.

SPEAKER_00 16:47

And Betran acknowledges that the core concept for Zing was fantastic, and the underlying technology idea was perfectly sound. But the project failed to gain the necessary momentum in the market because it became suffocated by traditional big bank risk aversion.

SPEAKER_01 17:03

The specific fatal flaw they point to is honestly mind-blowing. When HSBC launched Zing, they forced even their existing bank customers to completely restart the KYC and AML checks.

SPEAKER_00 17:15

Yeah, that was a massive misstep.

SPEAKER_01 17:17

Just for the listener, KYZ is know your customer, and AML is anti-money laundering. It's that intense identity verification process where you upload your passport and prove your address.

SPEAKER_00 17:27

Consider the friction that creates, right? A customer has been banking faithfully with HSBC for 10 years. The bank already monitors their direct deposits, their mortgage, their daily spending.

SPEAKER_01 17:37

They know everything about them.

SPEAKER_00 17:38

Right. Then the bank invites them to download this sleek new app, and the very first screen asks the customer to take a photo of their driver's license and upload a recent utility bill, treating a 10-year loyal client like an unknown suspect off the street.

SPEAKER_01 17:53

Yeah, a user will just close the app and never open it again. It completely destroys the frictionless experience that fintechs are known for. It's the equivalent of buying a Ferrari to compete in a race, but insisting the driver keeps the emergency brake pulled up the entire time just to be safe.

SPEAKER_00 18:10

That risk-averse siloed mentality is exactly what legacy thinking is. Patran points out that speed to market is a matter of survival today. If internal bureaucracy means it takes 18 months to launch a new feature, consumer expectations will have already moved on by the time you deploy it.

SPEAKER_01 18:25

Which brings up a really fascinating internal policy at Toom. To ensure they don't slowly calcify and become the very legacy system they are trying to replace, Toom dedicates a full 25% of its research and development budget specifically to retiring its own technical debt.

SPEAKER_00 18:40

Technical debt is an unavoidable byproduct of writing software. It is the perfectly functional code you wrote three years ago that solved a problem at the time. But as newer, faster methodologies emerge, that old code becomes a bottleneck.

SPEAKER_01 18:54

So by constantly rewriting and pruning their own architecture, they maintain their agility.

SPEAKER_00 18:59

Exactly.

SPEAKER_01 19:00

They're essentially paying to sweep the dust out of the house every single day rather than sweeping it under the rug for 30 years until the floorboards rocked.

SPEAKER_00 19:08

That's a great way to frame it.

SPEAKER_01 19:09

But even with a spotless code base and a brilliant cloud architecture, Toom faces an incredibly steep uphill battle. The CEO openly admits their most significant challenge is industry credibility.

SPEAKER_00 19:21

Oh, absolutely.

SPEAKER_01 19:22

They are a relatively young, small tech company attempting to unseat massive, entrenched industry titans like Oracle, Taminos, and Finastra.

SPEAKER_00 19:31

And those legacy giants have cultivated deep, decades-long relationships with conservative bank boards. A bank executive knows that nobody ever gets fired for buying the established legacy software, even if it is slow and expensive.

SPEAKER_01 19:44

It is the ultimate David versus Goliath scenario.

SPEAKER_00 19:47

Yeah.

SPEAKER_01 19:47

Tooman is asking a century-old, hyper-cautious financial institution to trust a startup with their most critical asset. To go back to the medical analogy, they aren't selling a minor peripheral tool. They are selling the heart and lungs of the financial institution.

SPEAKER_00 20:02

Yeah, it's a huge leap of faith.

SPEAKER_01 20:04

It's like asking a conservative bank board to let a brilliant 25-year-old medical student perform open heart surgery instead of the seasoned 60-year-old chief of surgery simply because the student has invented a highly efficient new scalpel.

SPEAKER_00 20:18

The perceived risk is astronomical. But synthesizing the reality of the banking landscape today, the seasoned surgeon's operating room is fundamentally broken. The legacy mainframe is shutting down for five hours every night to process batches while the patient is literally on the table.

SPEAKER_01 20:34

Well, yeah.

SPEAKER_00 20:35

The architecture cannot support AI and it cannot underwrite small businesses efficiently. Bank executives are finally realizing that doing nothing or relying on outdated systems actually carries a far greater existential risk than partnering with an agile technology provider.

SPEAKER_01 20:50

The cost of inaction has finally eclipsed the fear of change. So what does this deep dive into the invisible plumbing of finance mean for you, the listener? The next time you open your banking app and notice a genuinely helpful, AI-driven insight about your spending habits, or if your small business applies for a line of credit and gets approved in minutes without submitting a single paper tax return.

SPEAKER_00 21:13

Which is life-changing for a business owner.

SPEAKER_01 21:15

Completely life-changing. Or if your banking data connects flawlessly to your accounting software through an API, you will know the mechanism behind it. You will know that deep down in the basement, the bank finally stopped patching the Jenga Tower and successfully swapped out the heart and lungs of their operation.

SPEAKER_00 21:31

They executed progressive modernization while keeping the institution alive and running.

SPEAKER_01 21:35

Exactly. But I want to leave you with one final thought to mull over, building on Miles Butran's point about bringing a fintech mindset to traditional banking. For the last decade, the dominant narrative in finance has been that fast, highly specialized fintech startups were going to completely disrupt and dismantle the slow dinosaur traditional banks.

SPEAKER_00 21:55

Right. That was the prevailing wisdom.

SPEAKER_01 21:57

But if technology providers like Tugum are quietly equipping these heavily regulated, massively capitalized traditional banks with the exact same cloud native architecture, the exact same API connectivity, and the exact same AI capabilities as the startups, what happens to the disruptors? If the dinosaurs suddenly learn how to sprint, will the concept of a separate, distinct fintech industry even need to exist 10 years from now? Something to consider the next time you're waiting three business days for a paper check to clear.