Mikhail Gromov, CTO of the Bir Ecosystem – Interview Series

Mikhail Gromov, CTO of the Bir Ecosystem, is a technology and financial services executive with more than two decades of experience building digital banking platforms, large-scale technology organizations, and integrated digital ecosystems. At Bir, he leads the technology and platform strategy across banking, payments, lending, identity, loyalty, and e-commerce, while also overseeing research and development around artificial intelligence adoption across the software development lifecycle. Previously, Gromov served as CEO of Yandex.Bank, where he led the launch of a digital bank within the broader Yandex ecosystem, and as CTO of DOM.RF, where he helped transform the organization into a cloud-native, API-first digital platform. Earlier in his career, he spent more than eight years at Sberbank, leading technology transformation across its international operations and managing distributed engineering organizations comprising more than 2,000 engineers.

Bir Ecosystem is an Azerbaijan-based digital ecosystem that brings banking, payments, e-commerce, loyalty, identity, and other everyday services together within an integrated platform. Launched publicly in 2025 under PASHA Holding, the ecosystem includes Birbank, the Birmarket e-commerce platform, the m10 digital wallet, and MilliÖn payment terminals, alongside partnerships with services including Trendyol and BakıKart. Bir describes itself as the first fully integrated banking, payments, and e-commerce ecosystem in the Caucasus and reports more than five million active users, giving it significant reach within Azerbaijan as it expands the underlying technology infrastructure connecting consumers, businesses, financial services, and commerce.

You’ve spent more than two decades helping transform traditional financial institutions into large-scale digital ecosystems and are now leading this process at the Bir ecosystem in Azerbaijan. Looking back across that journey, what lessons most shaped your approach to building digital financial infrastructure at national scale?

I’ve lived through several major waves of change in the industry over the past twenty years – the shift from branch-based banking to digital banking, then to platform businesses, and now to ecosystems powered by AI. If there’s one key lesson I’ve learned during that time, it’s that the real challenges you face when building out digital services are rarely technological.

In fact, every successful digital transformation I’ve been a part of has centred around the same three pillars.

The first is engineering culture and reliability. If you want to be people’s top-choice financial institution, reliability has to be non-negotiable. Customers may appreciate innovative products, but they trust organisations that are consistently available when they need them, and that trust, once lost, is very hard to rebuild.

The second is platform thinking. No matter how good your products are, they need to be built on reusable core platforms – otherwise every new product or partner integration requires you to start from scratch. Once those foundations are in place, you can scale efficiently, accelerate product launches, and onboard new partners far more quickly. For example, it took us a year to build out BirBonus, our loyalty programme, and BirID, our unified single sign-on solution, and integrate them into our digital bank BirBank – because we were laying the platform foundations at the same time. Onboarding our second partner, Yango, however, took just one month. The same logic now applies to AI. Just as we invested in platform foundations that made future integrations dramatically faster, we’re now building the AI platform layer that will allow every product and service to be AI-native by design, rather than retrofitting AI into existing systems.

The third pillar is customer-centric product development. Customers expect digital services to cover every aspect of their financial lives – which means financial services can’t just be limited to a bank’s own channels. They need to integrate seamlessly into partner ecosystems, marketplaces, mobility services, and other customer journeys. Finance should work as invisible transactional and lending rails that help customers achieve those goals in just a few clicks.

The Bir Ecosystem has grown to more than 5 million users in a relatively short time, representing over half of Azerbaijan’s population. What factors allowed Azerbaijan to accelerate digital adoption so quickly compared to many other emerging markets?

First, the country already had strong macroeconomic fundamentals to build on. Internet penetration sits close to 90%, mobile penetration exceeds 100%, and more than 70% of web traffic comes through mobile devices. The ingredients for growth were already there. People had the devices; what hadn’t caught up yet was the supply of digital services to meet that demand.

Second, Azerbaijan has been extremely effective at learning fast and applying what works. Rather than investing heavily in research or building entirely new models from scratch, it has drawn on what’s already been proven in other markets – adopting successful models from abroad and instead doubling down on execution. That dramatically compresses the time it takes to build. From my own experience, what once took 5–7 years – including all the time needed to test and learn – now takes just 1–2 years. We’re building three to five times faster than we were a decade ago.

A third factor is a genuine culture of openness. In Azerbaijan, there’s a real willingness to listen, learn, and adopt new approaches regardless of where they come from. That might sound like a soft factor, but it makes an enormous practical difference – it means successful models can be implemented quickly and scaled efficiently, rather than getting stuck in internal resistance.

And finally, strong competition among local financial institutions has played its part too. Competition always has a way of accelerating everything.

Bir combines banking, payments, e-commerce, transportation, lending, identity, and loyalty systems into a unified platform. Why do you believe the “ecosystem model” is becoming increasingly important in financial services?

Customers don’t need financial products for their own sake. Nobody wakes up wanting a bank account – they wake up wanting to travel, shop, get around the city, pay their bills, and save money. Some of those journeys are fully digital, but others blend the digital and physical, like buying a home or a car.

While traditionally organisations have been structured vertically, customers live their lives horizontally. Ecosystems close that gap, meeting customer needs more seamlessly across the full breadth of their daily lives.

What makes the ecosystem model so attractive is that it enables organisations to participate in a much larger share of a customer’s life. From a business perspective, this creates a flywheel effect: driving cross-sell across services, deepening customer engagement, generating richer data that enables greater personalisation, and lowering customer acquisition costs.

As a result, financial institutions are increasingly facing a strategic choice: to either build ecosystems themselves or to become part of ecosystems built by others.

You previously helped launch Yandex Bank and worked on large-scale digital transformation efforts at Sberbank. How does building a fintech ecosystem in Azerbaijan differ from building one inside larger technology-driven economies?

The principles and strategies are largely the same, but the constraints are naturally different.

At a fundamental level, customers want the same things – simple, reliable, seamless financial services accessible in just a few clicks. That means many of the strategies for delivering those services transfer across markets relatively easily.

That said, every market has its own nuances. Infrastructure and connectivity in Azerbaijan, for example, vary widely across the country – while Baku is well served, in rural areas and mountainous regions connectivity can be less consistent. Attitudes toward digital financial services vary too. In some regions, cash still plays a more important role in everyday transactions, and digital services are seen as less personal than traditional interactions. Personal relationships, trust, and direct human contact remain important parts of doing business here, and that has to be built into how you design and deliver services.

The size and maturity of the market bring their own constraints as well. Unlike in larger, more technology-driven economies, we sometimes face a lag in local talent and competencies. But the flip side of being a smaller, earlier-stage market is that you carry less inertia. You can move quickly to adopt technologies that have already proven themselves globally, without the legacy systems and organisational complexity that hold larger markets back.

Bir has integrated products like BirID, BirPay, BirBonus, and BirCredit into a shared platform architecture. How important is digital identity infrastructure in enabling the future of embedded finance and online commerce?

Digital identity is one of the key enablers of embedded finance, but its value goes far beyond authentication. The real advantage comes when identity, payments, loyalty, and lending work together as a single integrated layer that any partner can embed into their channel – rather than connecting each service separately.

Our integration with Yango demonstrates this. Customers can use their loyalty balance to pay for taxi rides without having to create new accounts or enter additional information. Authentication happens through Face ID for App2App journeys and OTP for App2Web – and the entire experience becomes almost invisible.

To break it down, there are three levels to how digital identity creates value.

The first is scale. Identity is the foundation of reach – you can only serve customers you can recognise. In Azerbaijan, the digitally active population represents around 65–70% of the country – and with 5.1 million customers, we’re approaching that ceiling. We still have 20–30% growth potential, primarily by connecting new services and partners who bring in customers that don’t yet overlap with our existing base.

The second is engagement. Once you know your customer, you can draw them deeper into the ecosystem – across banking, e-commerce, transportation, and other areas – through services that work together seamlessly. Around 20% of our customers already use products across two or more verticals or partners, made possible by a single login, a shared loyalty programme, and built-in payment mechanisms. As more customers experience that connected journey, engagement compounds: more usage generates more data, better personalisation, and more reasons to stay within the ecosystem. We believe we can increase that cross-ecosystem usage by 1.5 to 2 times, with one key driver being moving customers from BirBank – which has our highest daily and monthly active users – into other areas like BirMarket.

The third is monetisation. Once you have an identified, engaged customer base, you can offer relevant products and services at exactly the right moment with very low acquisition costs. Identity is what makes that possible: you already know the customer, you understand their behaviour, and you don’t need to win them over from scratch each time.

This is why I believe digital identity is becoming one of the foundational layers of the future digital economy. Payments and loyalty can be copied, but identity is the connective tissue that holds the ecosystem together – and whoever owns the customer relationship can offer new services at the lowest possible acquisition cost.

Azerbaijan is still transitioning from a largely cash-based economy while mobile banking and digital payments continue to scale rapidly. What are the biggest infrastructure or consumer behavior challenges in accelerating that shift?

There are several challenges we’re up against in accelerating the country’s shift to digital payments. On the infrastructure side, despite Azerbaijan’s strong internet penetration, there are still areas – namely rural areas and mountainous regions – where connectivity is less reliable. Digital adoption always develops at different speeds across different parts of a country, and that uneven access remains an obstacle to full-scale adoption at this stage.

On a cultural level, cash is also deeply embedded in certain sectors of the economy, particularly in agriculture and small businesses. People are accustomed to using cash because it has historically been the simplest and most familiar payment method.

At the same time, customer behaviour is changing rapidly. Each year, more people are discovering the convenience of mobile banking, instant payments, and digital commerce. As they become more familiar with digital services and see first-hand how much simpler and more beneficial they can be, trust naturally grows and habits change. Our growth story is a testament to that.

Bir is deploying AI across customer operations, financial services, and internal engineering workflows. Where are you currently seeing the most practical real-world impact from AI adoption inside financial institutions?

The two areas where we’re seeing the most real-world impact from AI are customer support and product development.

At BirBank, AI now resolves 47% of customer requests without any human involvement, and is driving significant cost savings – every percentage point of autonomous handling translates to roughly 2% in Full-Time Equivalent (FTE) savings. AI customer support agents are also generating additional revenue, with conversion rates to product purchases already sitting in the 2–6% range.

On the product development side, AI adoption in software development has gone from around 50% in 2023 to over 90% by 2025. At Bir, we’ve seen features that previously took one to two weeks delivered in two hours – a 20x to 50x acceleration that fundamentally changes how we think about roadmap and resourcing.

Beyond these two areas, more mature organisations are making a strategic shift from AI assistants built for employees to AI assistants built for clients – enabling conversational banking, autonomous customer journeys, and more personalised marketplace experiences.

You’ve experimented with AI-assisted development tools like Cursor and Postman AI while leading large engineering organizations. How do you see AI reshaping software development and product delivery inside banks over the next five years?

Having led large engineering organizations and experimented hands-on with tools like Cursor for code review and Postman AI for writing automated tests, the shift I’ve seen is fundamental. AI is becoming a team member.

We’re already moving beyond copilot mode, toward multi-agent environments that power a fully AI-native Production Development Life Cycle (PDLC). The classic nine-role product team is giving way to something leaner. A “pit stop” model built around two or three roles, Product Owner, Tech Lead, and Quality Assurance Lead. This drives faster experimentation, tighter feedback loops, and the ability to tailor entire channels or mobile experiences to specific customer cohorts at a depth that was previously not feasible.

We’re also seeing product development become accessible to an entirely new range of people. For example, my daughter built a Minimal Viable Product (MVP) of her school website in twenty minutes, something that just five years ago would be unthinkable. Every time development has moved up an abstraction layer – from Assembler to C++, then to Python – the barrier to creation dropped, and prompts are the next step in that progression.

Yet we also need to be mindful of the risks associated with using these tools. Delegating code to AI agents risks eroding the system intuition engineers build by writing it themselves. There is, however, an underappreciated upside: AI-native Software Development Life Cycle (SDLC) produces documented solution architecture as a natural output, something human teams often don’t have time for. The token cost curve will also force a reckoning on where human judgment remains non-negotiable, likely pulling certain critical design decisions back into human hands.

Across your career, you’ve led cloud-native modernization efforts, DevOps transformations, and infrastructure upgrades supporting millions of users. What separates organizations that successfully modernize from those that remain trapped by legacy systems?

I think the fundamental mistake that many organisations make on this front is viewing modernisation as a technology problem. From my experience, modernisation is actually an organisational problem disguised as a technology problem. Companies often think that modernisation means moving to the cloud, replacing legacy systems or introducing the latest tech tools. While these things are certainly all important, they are not the primary reason behind what makes some transformations succeed and others fail.

On the organisational level, successful organisations modernise methodically, step-by-step. They focus on business outcomes, rather than implementing technology for technology’s sake. They invest in platforms that can be reused across many products, and – most importantly – they invest in people.

One of the biggest differences I see is culture. High-performing organizations think in terms of “we”, not “I”. Teams focus on customer outcomes and company goals instead of optimizing their own silo.

Another important factor is people – specifically, what I’d call T-shaped leaders. I often use the analogy of a professional singer: they may specialise in one style, but they still understand classical, pop, rock, and jazz. Modern leaders are the same. Deep expertise in one area is essential, but the ability to connect business, technology, operations, and customer experience into a single coherent vision is what makes the difference.

Organisations that struggle with modernisation tend to do the opposite. They focus narrowly on technology replacement, underestimate the culture change required, and treat transformation as a one-time project rather than an ongoing capability.

One lesson has stayed with me throughout my career, which is that legacy systems are often easier to replace than legacy ways of working.

Looking ahead, do you believe banking will increasingly disappear into the background as financial services become fully embedded into digital ecosystems, marketplaces, mobility platforms, and everyday consumer experiences?

I don’t think banking will disappear, but it will evolve from its current form.

If we look at the history of technology, we moved from low-level programming languages to high-level languages, and now increasingly to prompts written in natural language. The underlying technology did not disappear, it just became abstracted and easier to consume.

We’re witnessing a similar evolution in financial services. We moved from offline banking to online banking, and now we’re moving toward embedded finance, where financial products become part of broader customer journeys.

Today, it’s relatively easy for a fintech to become a customer’s third-choice bank – one people use simply to get cashback or discounts on a marketplace. Becoming a second-choice bank that covers most daily financial needs is harder. But becoming a customer’s primary financial institution is a different challenge entirely.

To reach that position, customers must trust you with their salaries, savings, investments, and long-term financial future. In my view, that level of trust is built over decades through stability, reliability, and the ability to navigate multiple economic cycles and crises.

The customer experience will change profoundly, but trust will remain one of the most valuable assets in the industry, and that isn’t something that will disappear into the background.

Thank you for the great interview, readers who wish to learn more should visit Bir Ecosystem.