Enterprises often hesitate to modernize legacy systems due to risk, complexity and budget concerns. In a Q&A with The AI Innovator, Stefan Aulbach, technology fellow at Deloitte Consulting, outlines how the company is leveraging generative AI to help its clients accelerate application modernization, enhance DevOps, reduce transformation risk and tie technology upgrades to long-term business value.
The AI Innovator: What are the most common challenges clients face when modernizing legacy applications, and how does Deloitte help them overcome these issues?
Stefan Aulbach: One of the biggest challenges to modernization is overall risk aversion because these legacy systems are typically still considered reliable. Another challenge is their size, as legacy systems have ballooned to include extensive code, data and interdependencies, leading to increased complexity.
Moreover, outdated and unclean data can lead to data quality, security or compliance issues during the modernization process. A lack of comprehensive and up-to-date documentation about the system can create challenges for the modernization team, as it strives to understand the existing system architecture and the business rules and overall functionality of the system. Finally, budget constraints tend to complicate modernization efforts for most organizations.
In what ways is Deloitte using AI to accelerate application modernization for clients?
Deloitte employs generative AI to better understand legacy systems, allowing us to completely and quickly review legacy code without needing to know the programming language concepts. AI is also used to spot repetitive patterns, thereby helping to increase automation by applying ‘refactoring recipes’ on a large scale.
Additionally, Deloitte has been using gen AI to extract the relevant business logic out of the legacy code, focusing on the key aspects of it instead of the boilerplate code. AI can further help in the generation of user stories, as well as in advancing towards autonomous coding using tools such as copilots or agentic programming IDEs.
Deloitte is also using AI to enhance DevOps capabilities, thereby accelerating infrastructure build-out, deployment automation, and QA processes.
Together, these efforts help to unlock the intellectual property hidden in millions of lines of legacy source code and terabytes of data – in addition to automating repetitive tasks, improving system reliability, and providing insights to help facilitate better and faster decision-making.
How does Deloitte help clients identify which applications to retire, refactor or rebuild from scratch?
Deloitte assists clients in identifying which applications to retire, refactor or rebuild from scratch through a comprehensive IT landscape assessment using our TruNorth platform. This tool helps clients understand their applications, the interconnections between systems, their infrastructure and, most importantly, overall condition.
The status of a client’s applications is determined by two measurements of fitness: their technological fit, such as maintenance costs and adherence to coding standards, and their business fit, which includes time to market and strategic importance. Part of this assessment also involves understanding the main drivers of modernization.
Once stakeholders have aligned on priorities, we create a shared vision and strategy, all leading to specific modernization journeys for specific applications. Each modernization journey may look different, but TruNorth helps to ensure it aligns with the agreed upon business case. Clients can also compare multiple journey options, along with their impact on cost, business continuity and alignment with modernization drivers.
How do you help clients future-proof their application ecosystems – ensuring they don’t just modernize for today, but build for tomorrow?
While it depends on the selected modernization journey, Deloitte helps clients future-proof application systems by integrating our solutions for language transformation, not only from a technical aspect, but also from a DevOps aspect, covering the full software development lifecycle.
For example, for rewrite, we support any target technology that the client wants, ranging from classic web applications to agentic applications. Any rewrite journey is aligned with evolving customer expectations. During the journey planning process, our Deloitte architects recommend and build a future-proof architecture alongside our clients’ architects. This effort is undertaken in early phases, so we can feed the planning back into TruNorth to see the various impacts on the modernization journey.
We also put systems in place to help them succeed well into the future, such as identifying coding standards at the outset of the project and putting processes and rules in place to audit and monitor for retention periods of materials.
What role does Deloitte play in helping clients select the right platforms, cloud providers and tools for modernization?
In assisting clients with selecting the appropriate tools and platforms, Deloitte first assesses and adapts to the client’s requirements. With this client-centered approach, we work with our broad landscape of technology alliances to help select the right technologies and bring the best people to the team to successfully deliver the projects and address any unforeseen challenges.
Skipping 50 years of technology and going from any of 10 mainframe languages (including COBOL and assembler/TPF) to cloud-native can be challenging for any workforce, so we also help our clients’ talent to develop and grow on-the-go while delivering on the modernization journey.
How does Deloitte help clients rethink their operating model and talent strategy to sustain innovation after modernization?
Rethinking the operating model is an integral part from the beginning of the journey. During the assessment of the application landscape, we will also review and refine the future operating model, together with the client. The client knows their system the best, so it is critical to have them in the conversation and to continuously improve throughout the process.
Many clients fear disruption to core operations – how do you help mitigate transformation risk while keeping critical systems running?
To mitigate transformation risk while keeping critical systems running, we take a repeatable and incremental approach.
We also look to minimize disruption by using selective refactoring – such as batch, UI/UX, or selected functionality. A high degree of automation and a test-driven approach are critical, ensuring that the future state still adheres to the behavior of the legacy system by moving existing test suites from legacy tech to the modernized tech. Finally, we use gen AI to generate test cases and test them against the legacy system and the new system.
How does Deloitte help clients connect modernization to broader strategic goals like improving customer experience, driving efficiency or enabling growth?
Once we ‘crack the mainframe,’ the possibilities become endless – new opportunities naturally arise by leaving a closed ecosystem with a stringent operating environment behind. Mainframes traditionally run a lot of workloads in batch, but customers today really want real-time insights. We help our clients in transforming to a real-time business, either as part of a modernization journey or as a totally separate journey.
Can you share a story that highlights Deloitte’s approach to delivering tangible business value through modernization?
One notable story that embodies Deloitte’s approach to using modernization to drive business value is our work for a major U.S. airline to improve their overall customer experience.
In order to do this, a broad modernization of aging core systems was needed. After many discovery sessions and breaking down our plan into three phases – strategy, mobilization and execution – it only took a few weeks for our client to gain insights into their entire passenger service mainframe using our mainframe modernization tool.
From here, they were able to decompose the mainframe and use gen AI to quickly extract and translate rules. As a result, the airline’s recovery costs will shrink significantly, and they now have the power to use gen AI chatbots in assisting both the customer to get answers to their questions and gate agents to deliver an automated service.
Be First to Comment