TBC Bank, the largest banking group in the Caucasus region with nearly 40% market share in Georgia, set out to become a digital-first leader while many competitors still relied on legacy systems. But with on-premises SQL Server instances, month-long reporting cycles and no platform for advanced analytics, the bank needed a modern data foundation to unlock a new way of working. After building its Lakehouse on the Databricks Platform and adopting Lakebase and Databricks Apps, TBC now powers self-service analytics and AI-driven applications that are changing how teams access data and make decisions.
On-premises tools could not keep pace with a digital-first ambition
TBC Bank operates across Georgia and Uzbekistan, serving millions of customers across retail, corporate, wealth management and MSME banking. Listed on the London Stock Exchange and recognized as one of the most data-mature financial institutions in Eastern Europe, the bank views technology as central to its competitive position.
Four years ago, TBC Bank began a major effort to standardize and modernize its data infrastructure, moving beyond on-premises SQL servers and slow reporting workflows to support advanced analytics, self-service access and AI-driven applications.
However, operational limitations made that transformation urgent.
Reporting inefficiency:
A single analysis could take up to a month, requiring multiple teams to work across disconnected datasets to identify questions, locate the right data and produce a usable report.
Delayed model deployment:
Credit risk scoring models took months to deploy, slowing loan disbursement in a market where speed to decision directly impacts profitability. Specifically, it took 14 weeks and now we are down to two days, including model risk validation process.
Transactional data constraints:
As teams sought to embed governed data into operating systems and customer-facing applications, the bank had no way to serve transactional workloads without spinning up standalone databases, duplicating security controls and increasing operational overhead.
“Without proper investment in data platforms from the very beginning of our transformation, we would never be able to achieve our goals,” said Konstantine Khoperia, Chief Data and AI Officer at TBC Bank. “We needed to build from scratch, in the cloud, with the right foundation.”
Lakebase bridges governed data and operational applications
TBC Bank built its Lakehouse on the Databricks Platform from the ground up. The team migrated all workloads to the cloud, adopted Unity Catalog as the governance layer and began consolidating querying and metadata management into a single environment.
With governed data in the Lakehouse, the team added Lakebase as the operational database layer, serving as a bridge between curated datasets and applications that needed fast, reliable reads and writes. For a team of just four DevOps engineers, the appeal was practical. “We deployed Lakebase in a couple of hours,” said Archili Janjibukhashvili, Head of Data Engineering. “For data engineers, this environment is completely familiar. There is no ramp-up, just building.”
The team delivered two early use cases, both built using Databricks Apps:
Reverse ETL that pushes governed data back into operating systems
Web-based AI chatbot with Lakebase as the transactional backend
Databricks Apps has become a catalyst across the organization.
With Lakebase, the data governance team can roll back tables to a previous state through a self-service application, rather than filing a request.
In the risk department, nontechnical users upload customer lists, select from multiple ML models and receive scored results almost instantly. Previously, this workflow required emails, file exchanges and days of data scientist time.
Meanwhile, Genie has emerged as a core self-service analytics capability. More than 600 users regularly query governed data through Genie, and adoption accelerated as deep research enabled business leaders to ask complex strategic questions and receive synthesized reports in minutes.
From month-long analyses to minutes, with AI augmenting every team
The shift in how TBC Bank operates has been immediate and measurable. By embedding governed data into operational workflows and enabling self-service analytics through Lakebase and Genie, the bank accelerated decision-making, streamlined model deployment and expanded AI adoption across the organization.
Faster analysis and decision-making
Analyses that once took a month now take minutes through Genie, and one of the senior leaders described uncovering more about their own business unit in three days of using Genie than in eighteen months of traditional reporting.
Credit risk scoring models that previously required months to deploy are now live in days through AutoML, directly accelerating cash loan disbursement, one of the bank’s highest-value business lines.
Workforce efficiency through self-service
Self-service tools have reshaped the workforce. TBC Bank was able to allocate data analysts to higher-value tasks reduced its data analyst team by 50% and reallocated their time to higher-value tasks as AI-powered capabilities absorbed routine analytical work.
“When you think about it in terms of savings and productivity, the numbers are significant,” said Konstantine. “But the harder thing to measure is what it means when someone gets an answer in seconds instead of three days. People feel it, and they love it.”
Adoption and scale across the organization
Today, roughly 20% of headquarters employees use Genie, and adoption is climbing. TBC Bank expects at least half of its 3,000-plus headquarters staff to be daily users by the end of the year.
“We spent years building a governed lakehouse,” said Konstantine. “Now Lakebase and Genie are putting trusted data in everyone’s hands, and decisions are moving so much faster.”