How to use AI in Application Modernization
Application modernization is often time-intensive and costly, but SUE’s latest TechTalk explores how AI can revolutionize the process. Experts Bob Merkus and Jeroen van Nieuwenhuizen discussed how Large Language Models (LLMs), Retrieval Augmented Generation (RAG), and Agentic AI can streamline modernization, tackling legacy infrastructure and skills gaps.
Recap TechTalk november 2024: How to use AI in Application Modernization
Application modernization efforts have historically been time-consuming and costly. But with new technologies, new possibilities arise. At SUE, we are innovators at heart. So when Generative AI had its breakthrough, we immediately were curious how we could apply this and help our customers innovate faster in their business. One way we did this was by looking at the use of AI in Application Modernization projects.
During our latest TechTalk, our AI and Application Modernization experts, Bob Merkus and Jeroen van Nieuwenhuizen, shared their vision, expertise and experience.
Common Challenges of Application Modernization
Jeroen started the talk by sharing common challenges he encounters at many organizations. There are multiple, but often these are included:
- Legacy infrastructure dependencies
- Skills gap
- Complexity of legacy code
- Risk of business disruption.
Often, these challenges can be solved by a lot of people that have a lot of time and a lot of money. According to Gartner, typical modernization projects last 16 months and cost about 1.5 million dollars. More than a quarter of the projects take two years or more and 93% of the respondents of a recent survey characterized their modernization experience as “extremely or somewhat challenging.”
So when Application Modernization efforts are time-consuming and costly, how can AI help solve these challenges? That’s what Bob talked about.
The Impact of AI on Application Modernization
In this talk, Bob provided an insightful overview of AI’s role in modernizing applications. After an overview of the AI landscape, he then delved into Large Language Models (LLMs) and Retrieval Augmented Generation (RAG), explaining how these technologies can significantly impact application modernization by enhancing information retrieval and providing relevant insights.
He then introduced the concept of Agentic AI approaches—specialized actors powered by LLMs that can interact dynamically with their environment by accessing information and performing pre-defined actions through tool calls. This agentic capability, he noted, allows AI to support more efficient application migration and modernization processes.
To conclude, Bob introduced Re:App, a new AI-driven tool from SUE, designed to streamline the replatforming of applications to cloud-native environments. A live demo showcased its potential to simplify and accelerate the replatforming process.
