Upskilling One of Central and Eastern Europe’s Leading Banks for the Databricks Era

OTP Bank is Hungary’s largest bank and one of Central and Eastern Europe’s leading banking groups. In 2024, OTP launched a major data platform migration, rolling out Databricks across multiple business units as part of a strategic modernisation programme. Alongside this technical transition, OTP recognised that its data and development teams would need more than standard product training. They needed a learning experience as sophisticated as the platform they were adopting.

Datapao Education was selected to design and deliver a bespoke, end-to-end Databricks enablement programme tailored to OTP’s specific technical environment, compliance requirements, and diverse workforce. What began as a single-cohort pilot has grown into a multi-cohort, multi-team initiative spanning the entire organisation.

The Challenge

OTP had developed a highly customised data infrastructure including its own proprietary internal libraries, a custom Windows Subsystem for Linux (WSL) image, strict data access controls, and enterprise-grade software engineering standards that apply to every line of code written on the platform.

Teams worked within a highly regulated environment where access to data was strictly controlled, compliance requirements varied by team, and code quality standards were non-negotiable. Risk management and strategy teams, for instance, were subject to tighter reporting and governance rules than internal data science teams, meaning the same platform was used in fundamentally different ways across the organisation.

Off-the-shelf training simply could not address these realities. Standard Databricks learning content provides a strong foundation in the platform, but it cannot by itself replicate OTP’s highly customised development environment, tooling and engineering standards. At OTP, developers are required to write code locally in VSCode, manage it through Git, test it with pytest, and deploy it with Azure DevOps, all within OTP’s bespoke toolchain and according to its own coding standards and documentation conventions. This is the way of the enterprise grade standard adoption streamlined with OTP’s banking standards.

The bank also faced a diversity challenge: the teams that needed Databricks training ranged from senior SQL analysts and business analysts with no software engineering background, to junior developers fresh from university, to mid-level managers who needed strategic fluency in the platform. Designing a single curriculum that could serve all of these learners (without pulling them away from their day jobs for weeks at a time) required a fundamentally different approach.

“We needed our people to work with Databricks the way we actually use it at OTP, not how it works in a textbook environment. That meant training in our own infrastructure, with our own tooling, on data that felt real.” (OTP lead)

The Solution: A Fully Custom, Cohort-Based Programme

Datapao Education worked closely with OTP to co-design a 10-week programme built from the ground up around OTP’s technical environment and learning objectives. The programme blends live instructor-led workshops, self-paced Databricks Academy modules, and a capstone pet project, all anchored in OTP’s actual systems. The 10 weeks duration balances the programme components and aligned with the participants daily workloads.

Rather than starting with a large-scale rollout, the two teams took an iterative approach. Together they designed a pilot programme that would test both the content and the learning model, with OTP contributing standardised coding templates and documentation that were embedded directly into the training materials, ensuring that participants encountered the same conventions in the classroom that they would use in production.

The pilot proved successful enough that OTP expanded the programme significantly. What started as a single cohort grew into ten additional cohorts for the following year, ultimately reaching approximately 150 employees across multiple business units, including the Data Science Technical Tribe, Retail Analytics, and Strategy & Analytics teams.

Programme Structure

The programme unfolds over 10 weeks per cohort, structured around three live contact days and supported by self-paced learning and regular office hours in between. The workload on the individuals is between 15-20 days per attendee, based on their professional background. 

Preparation Phase

The journey begins before the first live session, the Incubation Day. Each participant receives guidance on prerequisite knowledge: Python and SQL fundamentals, Databricks account setup, and preparation materials. OTP developed its own Python training that can precede the Incubation Programme for those with no Python experience. This phase establishes a common baseline and surfaces gaps early, allowing Datapao to begin personalising the experience before the cohort even enters the room.

Incubation Day

The first live workshop introduces the Databricks platform and the foundational concepts behind modern data architectures: Apache Spark, Delta Lake, and Unity Catalog. From the outset, the emphasis is on interaction and practical exploration rather than lecture. Participants leave with a working mental model of the platform and a clear sense of the learning journey ahead.

Self-Paced Learning with Personalised Paths

Following the Incubation Day, participants complete a survey based on which Datapao recommends a mix of Databricks Academy courses selected specifically for their knowledge level and role. 

The architecture of the programme is deliberately layered. The first half provides overall Databricks knowledge and includes both content tailored to OTP and some standard Databricks Academy content, personalised to each individual’s assessed starting point. The second half is fully custom and OTP-specific, focusing entirely on how Databricks is used within the enterprise environment of the bank. This approach creates a unique blend of depth, relevance, and scalability.

Optimisation Workshop

The second live day focuses on Databricks storage and performance optimisation, a deep dive into the technicalities of Delta Lake tables and trade-offs between cost and performance. It opens with an integrated FinOps session delivered by an OTP internal expert, covering how Databricks and cloud costs are tracked, attributed, and controlled within the bank. This module was not part of the original curriculum: it emerged from participant feedback mid-programme and was rapidly incorporated as a joint effort between Datapao and OTP, a concrete example of the programme’s adaptability.

Ways of Working Workshop

The final live day is the most OTP-specific of all. Participants work through development in WSL, OTP’s repository structure, branch-based development workflow, unit testing with pytest, and deployment using Azure DevOps. By the end of the session, they have cloned a repository, written a unit test, and submitted a pull request, operating within the same engineering model used across OTP’s development teams.

“The goal was to take someone who understands data but may never have used Databricks before, and guide them through the entire lifecycle, until they can work like an enterprise software developer on the platform.”

Office Hours

Between each live workshop, two office hour slots of two hours each give participants a dedicated time to ask questions, troubleshoot issues, and get help configuring their local development environment.

The Pet Project

The programme concludes with a Pet Project: a real-world task drawn from OTP’s actual reporting obligations, requiring participants to apply everything they have learned in a production-like setting (clone a repo, implement the solution, write tests and open a pull request).

Training in OTP’s Own Environment

All training takes place within OTP’s own Databricks workspace, using synthetic financial transaction data designed to closely resemble the bank’s real datasets without exposing any sensitive customer information. Participants use OTP’s internal libraries and follow the same coding standards and documentation conventions they will use in their daily work.

This meant Datapao trainers also had to become fluent in OTP’s own technical stack. 

Personalised Learning Paths

Datapao provided a complex enablement program. This approach allows the programme to serve a broad range of prior experience by utilising existing training material on Databricks Academy for generic foundational content and applying Datapao’s training development expertise where it adds the biggest value: customised training on client-specific use cases and data with content tailored to mimic how Databricks is used in practice within the organisation.

Designed for Four Distinct Learner Profiles

One of the central design challenges was the heterogeneity of each cohort. OTP’s data teams bring together professionals with radically different backgrounds  (and often, different levels of seniority) within the same training group. Datapao addressed this by designing for four distinct learner personas:

Because cohorts were mixed in practice (bringing together participants from multiple personas) Datapao trainers drew on their experience to adapt delivery in real time. Peer coding, group exercises, and LEGO-based collaborative activities were used to bridge experience gaps and create learning moments that worked for the whole room.

A True Collaborative Partnership

The programme was not simply delivered by Datapao to OTP. It was built together. 

OTP contributed standardised coding templates and documentation that became part of the curriculum. An OTP internal expert delivered the FinOps session. 

When it became clear mid-programme that the participants had varying levels of Python skills, both teams responded quickly: OTP built and launched an internal Python course; Datapao adjusted its onboarding expectations and office hour focus accordingly. This kind of rapid, joint response is what has allowed the programme to continually improve across cohorts.

“We responded to needs as they emerged. When participants needed something we hadn’t originally planned for, we found a way to deliver it, often together with OTP’s own experts.” (Datapao’s technical instructor)

Feedback is collected after every live session and at the end of each cohort, giving Datapao real-time data to refine both content and delivery style. The curriculum itself has remained stable; what evolves is how experienced instructors present and adapt it for each group.

The objective was never simply to teach Databricks. It was to help OTP’s people become productive in the bank’s data ecosystem as quickly and effectively as possible.

Results

Following a successful pilot cohort in 2025, OTP expanded the programme to ten additional cohorts for 2026, training over 150 professionals across multiple business units. Demand grew organically: additional teams joined the programme after seeing early results, and the initiative quickly expanded beyond its original audience.

Organisational Scale

• Expansion from one pilot cohort to ten additional cohorts
• Approximately 150 participants trained across multiple business functions and departments
• Adoption driven by demand, additional teams joined after seeing early results

Technical Readiness

• Faster onboarding into OTP’s Databricks ecosystem from day one
• Consistent development practices, coding standards, and documentation across teams
• Participants working confidently with Git, unit testing, and CI/CD in production environments
• Greater confidence operating within OTP’s full enterprise development toolchain
• OTP’s senior engineering team remained focused on building the new platform, while Datapao took ownership of onboarding and enablement, accelerating adoption without slowing platform development.

Leadership Alignment

• Active participation from management at department head and senior manager level
• Dedicated leadership cohort ensuring decision-maker alignment with the platform transition
• Stronger alignment between technical and business stakeholders across the data organisation

Sustainable Adoption

• A repeatable, cohort-based learning framework that OTP continues to run
• Continuous feedback and improvement process embedded in every programme cycle
• Long-term capability development across current staff and incoming talent

Looking Ahead

The OTP Databricks enablement programme demonstrates that successful platform adoption requires more than technology training. Organisations need learning experiences that reflect their own environment, processes, culture, and business realities, and partners who are willing to build those experiences collaboratively rather than deliver off-the-shelf content.

By combining deep Databricks expertise with close collaboration, hands-on learning, and genuine responsiveness to emerging needs, Datapao helped OTP transform platform adoption into sustainable capability development.

The result was not simply a trained workforce. It was a growing community of Databricks practitioners (spanning SQL developers, analysts, engineers, junior talent, and management) equipped to operate effectively within one of the most sophisticated enterprise data environments in the region.

About Datapao Education

Datapao Education designs and delivers bespoke data engineering and analytics training programmes for enterprises. Specialising in Databricks, cloud data platforms, and modern data workflows, Datapao works alongside clients to build programmes that fit their technology, their culture, and their people.

Planning a Databricks rollout or looking to build stronger in-house data capabilities? Talk to Datapao Education about a training programme designed around your teams, tools and real working environment.

Talk to our Education team: https://datapao.com/contact/