Bank of Georgia is a systemically important and leading financial institution in the country. It dominates the payments space and mobile banking, with strong retail and corporate banking franchises. In line with its digital strategy, the bank focuses on expanding its technology and data analytics capabilities to deliver more personalized solutions and seamless experiences for customers.

Today, 52% of all payment transactions in Georgia are processed through the Bank of Georgia. Each month, 1 million users actively engage with the bank’s core digital product, and half of them use the app daily. The bank also services around 44% of loans and 40% of deposits in the country.

Leadership at the bank believes the best way to build a modern and resilient company is by being technology-driven and customer-centric.

Here’s how Bank of Georgia navigated digital transformation and achieved full infrastructure visibility. 

Vaja Mantua — Deputy CIO, IT Operations. On digital transformation, impact, and necessary change

Digital transformation at Bank of Georgia

While many companies accelerated their digital strategies during the COVID-19 pandemic, we had already made significant progress by then. Today, we’re building on that foundation with unique features across our digital platforms.

Our ecosystem goes beyond traditional banking. For example, our platform includes a stock exchange feature where users can buy or sell shares with a single click. In the future, we plan to integrate a crypto exchange into our mobile banking app.

Our biggest challenges revolve around scalability, performance, security, availability, and streamlined business processes. In short, we deal with three main dimensions: people, processes, and platforms. 

People
Three years ago, we went through aggressive hiring, which revealed a few gaps — particularly in onboarding processes. We found that scaling a company 3–4 times over can impact service quality, which often hinges on people. That’s why we built an efficient onboarding process centered around mentorship, code review, and teamwork.

Processes
We transitioned from traditional waterfall processes to agile — not just as a workflow change but as a cultural shift. It pushed us to think bigger.

We introduced measurable tech KPIs, like APIs (Application Programming Interfaces). A culture of feedback is also growing — we believe listening to our employees and acting on input is key to success. A satisfied employee is a vital success factor.

Platforms
As our platforms became more complex and our user base grew, we needed a flexible, modern architecture. That meant microservices and lightweight processes that scale easily. 

What drives us to change 

Most companies respond to customer behavior; progressive ones shape it. We believe Bank of Georgia is in the latter group. That said, customer usage also pushes us to adapt and refine our platforms.

If you run a monolithic architecture, you’ll need a major overhaul every two years. That’s expensive and inefficient: when each process is performed in one program and platform, it is difficult to update them annually. That’s why we’ve adopted microservices — they're lightweight, scalable, and cloud-ready. In addition, this architecture will help us in our future migration to the cloud. We have announced that in five years, we will be a fully cloud-based organization, and all our platforms will run in the public cloud: Amazon, Azure or Google.

Microservices bring flexibility but also infrastructure complexity. To manage it, full visibility is critical to control scaling, process and architecture changes. Otherwise, they will not be able to bear the weight of their problems when moving to microservices, the cloud or wherever. APM solutions continuously assess system health and show how changes affect user experience and service quality. With full visibility, IT teams can simplify operations and accelerate digital transformation. 

How to choose an APM solution 

When searching for new tech solutions, we take two paths: internal research and trusted partner recommendations.

We have a dedicated team of architects in the IT organization for internal sourcing. They conduct research, find technical solutions for various tasks and make a first assessment. Occasionally, we choose solutions based on the advice of our local partners. We have about 6 partners whom we can trust.

Here’s our typical approach: ● Define your exact needs. ● List requirements and start exploring. ● Shortlist options that meet your criteria. ● Make a final decision.

It is important to clearly define roles and responsibilities in advance. Otherwise, you risk disappointment when tools don’t meet expectations.

We typically evaluate 4–5 major platforms, assess the project cost, and verify that it fits within the budget. Once a successful proof of concept is complete, we commit.

While choosing a monitoring tool, we asked: ● Who’s responsible for deployment? ● Who manages metrics and alerts?

Dynatrace stood out as the clear winner. In my opinion, it is the best application monitoring tool. The main reasons why we picked Dynatrace were the technical capabilities of the platform, the metrics it could provide compared to competitors, and the cost. Often, top-tier tools are expensive, but Dynatrace fit our budget without compromise. Their flexibility to adapt to our needs was a huge plus.

With 800 people in our IT department, having clearly defined roles and the right platform leads to great outcomes. 

David Lapiashvili — Head of Mobile & Web Development. On implementation and technical challenges

Dynatrace: The importance of implementation 

In 2017, we launched a new mobile app, which quickly gained traction, reaching 90,000 installs. But it came with challenges we weren’t prepared for. We handled issues manually, collecting data to identify service disruptions.

Customer satisfaction became our top priority, and that’s when Dynatrace entered the picture. Initially, we used it to monitor the Java-based back end of the mobile app. It proved invaluable in resolving problems efficiently.

Later, we migrated from a standard VMware setup to a VMWare + Kubernetes architecture. We anticipated issues with monitoring. But Dynatrace is effortless to work with new platforms and Kubernetes. Migrating licenses from VMware to Kubernetes was a smooth process.

The implementation took time, but Dynatrace’s documentation helped. We set up alert systems, tracked critical services, and defined business indicators. In about a month, we had dashboards, attribution models, and full monitoring in place.

Dynatrace remains our go-to for monitoring and troubleshooting. At the same time, we have a lot of KPI charts and graphs, small platforms, or non-critical processes. For this, we also use open-source tools like Prometheus.

Dynatrace measures SLA metrics too, helping us assess service stability — not just for us but also for integrated partners. 

How we defect issues 

We currently have two types of detection. The first is when customers are unaware of the issue, and Dynatrace identifies issues users aren’t even aware of. This lets us respond quickly — an enormous advantage.

The second is when we miss issues, or they stem from external factors (like vendor changes). This also creates certain fluctuations in our system and often affects the entire infrastructure. Dynatrace lags them in minutes, so we can fix them before they escalate.

Even when multiple systems fail at once, Dynatrace helps pinpoint the root cause, speeding up response time. It organizes issues chronologically and lets you drill down into any specific application.

We don’t use automatic rollback since most deployments include multiple components. Still, Dynatrace helps us track which service was deployed and where problems lie. 

The results we achieved 

When we started working with Dynatrace, it revealed issues we hadn’t previously detected. For example, we analyzed new call paths and exceptions. And also created new alerts for improved monitoring visibility.

Before Dynatrace, customers reported outages and disruptions. Now we get real-time alerts to understand where the issue is. As a result, customer satisfaction rose significantly. We now score 90 points for mobile applications and individual customers, more than 80 for business customers, and also more than 80 for individual web applications. This reflects major improvements in product quality and customer satisfaction.

We now measure downtime per system. Dynatrace helps us detect and fix root causes before users even notice them, dramatically reducing incident response times.

Bottom line: unscheduled downtime has been cut in half, which is a major win. The APM solution has significantly boosted service quality. We plan to expand coverage and monitor more of our infrastructure and critical systems. 

Final thoughts 

Our verdict: we highly recommend Dynatrace.
Monitoring every critical system with Dynatrace is a game-changer. The platform is packed with tools that simplify issue detection and resolution. We strongly recommend it to ensure the stability of key systems. 
Dynatrace goes beyond APM — it’s a full observability solution. We’re now exploring user monitoring, AppSec, and cloud monitoring features, and may expand our subscription. 
Before committing, we suggest building your own observability chain by connecting the frontend, backend, databases, and other components of a single service. This will help you fully understand what Dynatrace offers. 
Smaller companies often lack deep analysis tools. We hope our story shows how a solid monitoring platform can transform your operations and ensure software reliability.