Digital transformation in banking and AI in the banking sector are closely interconnected topics that are increasingly discussed in today’s technological landscape, where the financial industry is facing a pivotal crossroads. On one side, there is the need to innovate in order to respond to a market that is becoming ever more agile, digital, and regulated (think of RegTech and the AI Act).
On the other, there is the burden of decades-old technological infrastructures layered and consolidated that represent both the beating heart and the operational “cage” of many institutions.
Within the context of digital transformation in banking, these legacy systems serve as the operational core but also as a major barrier to innovation.
At Revelis, driven by the mindset captured in our tagline #artificialintelligenceinaction, we recently managed the migration of a critical banking procedure shaped by 30 years of continuous development. Originally written in legacy languages such as PL/I and COBOL , it was transformed into a modern microservices architecture based on Java/Spring Boot and Angular.
Below, we analyze a concrete case of digital transformation in banking, where we converted a monolithic system into a modern, scalable solution ensuring quality through SonarQube and achieving accurate data migration with AI-assisted ETL processes.
What is digital transformation in banking?
Let’s start by clarifying what digital transformation in banking actually means: it is the process of technological and organizational modernization through which financial institutions update systems, processes, and business models by integrating technologies such as artificial intelligence, cloud computing, and microservices.
This process enables banks to:
- improve operational efficiency
- reduce costs associated with legacy systems
- deliver faster, more personalized digital services
- ensure regulatory compliance (RegTech and AI Act)
- enhance data security and management
Digital transformation in banking is therefore a strategic step for competing in an increasingly digital and regulated market.
Examples of digital transformation in banking
Examples of digital transformation in banking include the adoption of new technologies to improve processes, services, and IT infrastructure.
Key examples include:
- migrating from legacy systems to microservices
- using AI for analytics and automation
- implementing chatbots and virtual assistants
- digitizing customer onboarding processes
- leveraging cloud computing for scalability and security
- integrating RegTech solutions for compliance
These initiatives enable banks to become more agile, efficient, and competitive in the digital marketplace.
The migration process in digital transformation
To better understand the approach taken, it is useful to look at how automation and human intervention were integrated at every stage of the transformation process.
Legacy systems: the weight of the past
Legacy banking software is not just “old” it is one of the main obstacles to digital transformation in banking. Over 30 years, multiple development teams maintained procedures written in PL/I and COBOL. While these languages were once pillars of enterprise computing, today they represent significant operational risks and are ill-suited for modern transformation needs because they:
- increase operational risk
- slow down innovation
- complicate integration with new technologies
The problem of technical debt
Technical debt is one of the biggest challenges in digital transformation. Thirty years of changes result in thousands of lines of code where business logic is often intertwined with outdated technical constraints. In many cases, original documentation has been lost or no longer reflects the reality of the code.
This creates a dangerous dependency on a small number of domain experts who understand these “black boxes.”
Migrating such a system does not simply mean “changing programming language” it means decoding three decades of business decisions, compliance rules, and complex calculation logic. This is where traditional approaches fail due to excessive costs and risks. This is also where Revelis’ vision comes into play.
Vibe coding and AI in banking transformation
As part of the transformation process, we adopted the Vibe Coding AI paradigm.
This approach enables:
- faster development
- improved code quality
- reduced dependency on legacy languages
What is vibe coding?
Vibe Coding is not just a trend it represents a fundamental shift in how developers interact with code. Instead of writing instructions line by line, developers engage with AI at a higher conceptual level. They communicate the “vibe” the intent, logic, and desired structure while the Large Language Model (LLM) handles the syntactic generation.
Prompt engineering as a programming language
Prompt engineering has become a key component in digital transformation. To extract logic from PL/I and COBOL, we used advanced prompt engineering techniques.
Rather than simply asking the AI to “translate,” we built structured prompting systems to guide deep reverse engineering:
- Contextualization: Providing the LLM with legacy language manuals to ensure accurate understanding
- Decomposition: Breaking large PL/I files into manageable logical blocks while preserving global context
- Iteration: Refining results through feedback loops, testing, and co-creation
Spec-driven methodology: the role of human validation
Human validation remains essential. The Spec-Driven approach allowed us to create an intermediate layer of functional “truth” before rewriting the system.
Generative technical-functional analysis
We used LLMs to analyze PL/I and COBOL source code and generate complete technical and functional documentation describing what the procedure does independent of the original language.
This analysis was not taken at face value. At Revelis, we see AI as an enhancer, not a replacement for human intelligence. Domain experts rigorously validated the output. Senior developers and business analysts ensured that the AI correctly interpreted even the most subtle business rules.
AI-Generated development plan
Once validated, we asked the AI to generate a detailed development plan, including:
- API definitions
- database structures
- integration logic between microservices
The AI effectively acted as a senior software architect, proposing the most efficient decomposition of the original monolith.
Data migration: AI-Assisted ETL processes
Data migration is one of the most critical aspects of digital transformation in banking. A banking system is nothing without its data. Thirty years of operations result in massive volumes of information stored in outdated formats or hierarchical/flat databases typical of mainframe environments.
As part of the transformation, this data must be migrated to modern relational databases.
We used LLMs to generate dedicated ETL (Extract, Transform, Load) procedures:
- Extraction: Understanding COBOL and PL/I record structures
- Transformation: Applying cleaning and normalization rules to ensure referential integrity
- Loading: Automating SQL script generation to populate new databases
This approach reduced database migration time by 80% and eliminated typical human errors in critical phases.
The new technology stack: microservices and modernization
Digital transformation requires modern architectures designed for scalability:
- Backend: Developed in Java with Spring Boot, breaking logic into independent microservices (e.g., customer data, accounts) to allow targeted updates
- Frontend: Angular replaced legacy character-based interfaces, significantly improving user experience and operational efficiency
Quality assurance with SonarQube
Code quality is crucial. AI-generated code introduces a key challenge: ensuring it is clean, secure, and free of vulnerabilities.
We integrated SonarQube into the development process to analyze every line of code for:
- Code Smells: ensuring readability and maintainability
- Vulnerabilities: critical for protecting sensitive banking data
- Technical Debt: avoiding long-term issues
This allowed us to objectively demonstrate that the new system is not only more modern but also superior in quality.
Efficiency and savings: the role of RegTech
Digital transformation is not just about technology it also involves regulatory compliance.
This approach ensured compliance with modern standards and the AI Act, while improving efficiency through automation of critical processes.
The AI Act introduces new requirements for transparency and governance. By documenting every step from reverse engineering to development planning we ensured full traceability and alignment with emerging regulatory expectations.
Conclusion: the future of digital transformation in banking
Compared to traditional manual migration, the benefits were clear:
- 50% reduction in overall development time
- 40% reduction in project costs
These results demonstrate that digital transformation in banking delivers tangible, measurable value.
Digital transformation is no longer optional, it is a necessity. The migration of PL/I and COBOL systems was not just a technical project; it proved that with the right methodology and a conscious use of AI, it is possible to unlock the hidden value of legacy systems.
Thanks to Vibe Coding, we turned IT archaeology into future-ready architecture. We demonstrated that:
- 30-year-old code can be understood and modernized rapidly
- AI, guided by domain experts, can produce reliable analyses and development plans
- modern technologies like Spring Boot, Angular, and microservices enable agile banking
- quality must always be measured and certified (SonarQube)
At Revelis, we continue to push the boundaries of innovation, bringing AI solutions across industries from healthcare to manufacturing and into the very core of banking systems.
If your organization is struggling with outdated systems that hinder growth, it’s time to change the vibe.
Want to start your digital transformation journey in banking? Get in touch with us.
Author: Massimiliano Ruffolo
Most frequently asked questions
Why is digital transformation important for banks?
Digital transformation in banking is essential to:
- remain competitive in the market
- reduce operational costs
- improve the customer experience
- adapt to regulations (such as the AI Act and RegTech)
Which technologies drive digital transformation in banking?
The key technologies behind digital transformation in banking include:
- artificial intelligence
- cloud computing
- microservices
- big data and analytics
- process automation (RPA)
What are the challenges of digital transformation in banking?
The main challenges of digital transformation in banking include:
- integration with legacy systems
- managing organizational change
- cybersecurity
- regulatory compliance
- shortage of digital skills
How much does digital transformation in banking cost?
The cost of digital transformation in banking varies depending on the complexity of existing systems, but it can be reduced through the use of artificial intelligence and gradual modernization strategies.
