By Monica Yadav \u00b7 June 25, 2026<\/p>\n<\/div><\/div><\/div><\/div><\/div>
Decades of business logic, process decisions, and hard-won rules are encoded inside your IBM i applications. The system works, it runs the business every day, reliably, without complaint. But the moment a change request lands, a compliance deadline looms, or a modernization project kicks off, one question surfaces faster than any other: does your team actually understand what is in there?<\/strong><\/p>\n<\/div><\/div><\/div> Key Takeaways<\/span><\/p>\n<\/div> The core problem: <\/b>decades of business logic live inside IBM i applications that documentation rarely captures fully, and the experts who understood them are retiring.<\/p>\n<\/div><\/li> The hidden cost: <\/b>developers spend 52 to 70% of their time understanding existing code, and discovery alone can consume the majority of a modernization budget.<\/p>\n<\/div><\/li> The solution: <\/b>DISCOVER automatically reads and maps your entire application (every program, dependency, and data path) and makes it explorable in plain language with AI.<\/p>\n<\/div><\/li><\/ul><\/div><\/div> Most shops will pause before answering honestly. The developer who wrote the order entry programs in RPG retired three years ago. The person who knows the nightly batch jobs lives in one city; the one who understands the EDI interfaces works in another. The business analyst who can explain what the pricing calculation does has never opened a CL script. Meanwhile, the COBOL modules that handle the general ledger posting have not been touched since 2011, and the two people who understood them deeply have long since moved on.<\/p>\n The complete picture of the application exists, every rule, every dependency, every calculation, but it lives in pieces: spread across job descriptions no one has updated, libraries no one has fully catalogued, and the memories of people who may or may not still be in the building. That is how every well-built, long-running IBM i application evolves. And right now, two forces are making that picture more important than ever to get a hold of.<\/p>\n<\/div> In Fortra’s 2025 IBM i Marketplace Survey<\/a>, IBM i skills ranked as the second biggest concern across the platform community, just behind cybersecurity, and it is actively displacing application modernization as organizations come to terms with what the retirement of key personnel<\/strong> actually means.<\/p>\n The challenge is straightforward: IBM i platforms contain decades of business logic that documentation rarely captures fully. As the engineers who built and maintained these systems retire, organizations face knowledge loss that makes modernization harder and riskier. The system keeps running. The understanding of it quietly thins.<\/p>\n Research confirms the wider pattern. Studies consistently show that developers spend between 52 and 70 percent of their time understanding existing source code, and that ratio is most pronounced on systems with the deepest history. The 2024 State of Developer Productivity report<\/a> found that gathering context and hunting for the right information ranked as the single biggest productivity leak engineering leaders hear about from their teams. On a core IBM i application three decades in the making, that search does not get easier with time.<\/p>\n<\/div> DISCOVER<\/a> is built to answer that core question, what is actually in your application<\/strong>, completely and automatically. It reads your application and maps the whole of it in a graphical view: every program, every dependency, every path a piece of data takes from a database file through to a screen<\/a>. Not from a document that went stale in 2014, but from the code as it stands today<\/strong>.<\/p>\n DISCOVER<\/a> works at a higher level of global application analysis, tracking how everything is hooked together, with dependencies on data and workflow explicitly mapped and diagrammed so they can be visually represented.<\/p>\n Check out this article: DISCOVER: Global Application Analysis With An AI Interface<\/a><\/p>\n AI-powered capabilities then make that map actionable for the whole team:<\/p>\n<\/div> Functional Tree powered by AI:<\/strong> Decades of programs rarely follow tidy naming conventions. Thirty programs might together handle order entry without the letters “ord” among them. DISCOVER<\/a> studies how they call and feed one another, recognises the business function they form as a whole, and automatically names and organises that branch of the map in plain business terms, so the functional structure of the application is immediately readable without any manual effort from the team<\/a>.<\/p>\n<\/div><\/li> Code explanation powered by AI:<\/strong> Open a program you have never seen and DISCOVER<\/a> tells you what it does, from a short summary an analyst or auditor can rely on, to a full technical account for the developer who has to change it.<\/p>\n<\/div><\/li> AI Assistant:<\/strong>\u00a0Ask in plain language, in English, French, or Spanish, where a field is calculated, what updates a file, or what a program is responsible for. The answer comes back in seconds, with no RPG or SQL knowledge required.<\/p>\n<\/div><\/li> On leveraging the ARCAD repository to enhance AI usage<\/b><\/span>:<\/strong> One of the key enablers in making AI truly effective on legacy applications is the availability of a structured and trusted knowledge base. This is precisely where the ARCAD repository plays a critical role. By capturing and organizing the application\u2019s metadata, relationships, and business logic in a deterministic way, it provides a reliable context layer for AI-driven exploration. Instead of forcing AI models to interpret raw code in isolation\u2014an approach that is both costly in tokens and prone to approximation\u2014queries can be grounded in a curated, pre-analyzed repository. The most common event in the life of an IBM i application is a change request. A new field on a screen, a tweak to a calculation, one more rule for one kind of account. Before a line is written, the real work begins: what does this touch? Which programs read that file? Where does that value get recalculated on its way to the report?<\/p>\n The discovery and documentation phase alone can consume the majority of a modernization project’s budget<\/a> before a single line of new code is written. And it does not stop there. Technical debt acts as a multiplier on every other cost factor<\/a>, making undocumented, tightly coupled applications significantly more expensive to change over time.<\/p>\n<\/div> Among the most powerful use cases enabled by DISCOVER<\/a> is data lineage analysis. In complex IBM i environments, understanding how a piece of data flows across programs, files, and processes is often a major challenge. DISCOVER<\/a> reconstructs this lineage end-to-end, tracing how data is created, transformed, and consumed across the application landscape. Whether preparing for a regulatory audit, a modernization initiative, or a functional enhancement, having instant visibility into data lineage turns what used to be a lengthy investigative effort into a matter of minutes.<\/span><\/p>\n<\/div> A developer in his first month on the platform can follow the same dependency chain, ask the same question in plain language, and get the same answer an expert would. DISCOVER<\/a> is used not only by RPG developers but by Java developers trying to understand how an application is built, and how the code is structured. The knowledge that once sat with a handful of people becomes something the whole team can reach.<\/p>\n<\/div> DISCOVER<\/a> fundamentally changes who can understand and interact with legacy applications. Traditionally, decoding RPG-based systems required highly specialized technical expertise, often concentrated in a shrinking pool of experienced developers. With DISCOVER<\/a>, this barrier is removed. Through intuitive visualizations, business-oriented navigation, and structured insights, non-technical analysts can quickly grasp how the application works, how processes are linked, and where key logic resides. This democratization of knowledge accelerates onboarding, improves collaboration between IT and business teams, and reduces reliance on scarce technical resources\u2014allowing organizations to unlock the full value of their existing systems without needing to rewrite them. <\/p>\n Listen to the Tech Talk: AI-driven application analysis with DISCOVER<\/a>, a short walk through how it maps a real IBM i application.<\/p>\n<\/div> When the business case calls for exposing IBM i logic through REST APIs, connecting to cloud services, or feeding data into AI models, knowing exactly what is there and what touches what is the work that makes integration go smoothly. The map DISCOVER<\/a> produces is the foundation that every downstream decision can stand on.<\/p>\n<\/div> IBM i modernization has always worked best when it starts from solid ground. With AI and machine learning now a top and growing concern across the IBM i community, teams are being asked to point new technology at systems they may only partially understand. The answer is not to slow down. It is to know the system completely first, so every next step is taken with confidence.<\/p>\n DISCOVER<\/a> makes the application readable. Not just to the people who built it, but to everyone who needs to work with it, change it, or build on top of it, starting today.<\/p>\n<\/div><\/div><\/div><\/div><\/div> Map your application in minutes and let your whole team explore it, in graphical view and plain language, right from the browser.<\/p>\n<\/div> Monica Yadav is a Solutions Architect at ARCAD Software. She has more than five years of experience in software quality assurance. She works closely with clients, developers, and product teams to understand technical requirements, define solution strategies, and support successful product implementations. Her expertise includes DISCOVER<\/a>, and ARCAD CodeChecker<\/a>. Monica also contributes to product demonstrations, training, documentation, and knowledge-sharing initiatives.<\/p>\n<\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/div><\/p>\n","protected":false},"excerpt":{"rendered":" Cut impact analysis from days to hours. DISCOVER auto-maps every program, dependency and data path across your IBM i application with AI.<\/p>\n","protected":false},"author":1,"featured_media":1434,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[50],"tags":[],"class_list":["post-1398","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog"],"yoast_head":"\nThe complete picture exists, but it lives in pieces<\/h2><\/div>
The problem: the knowledge gap is real and growing<\/h2><\/div>
Let the application show you the whole picture<\/h2><\/div>
<\/span>This not only <\/span>dramatically reduces token consumption<\/strong> and response latency, but also <\/span>significantly improves the accuracy and relevance of insights<\/strong>, anchoring AI reasoning in facts rather than assumptions.<\/span><\/span><\/p>\n<\/div><\/li><\/ul>The moment it pays off<\/h2><\/div>
Impact analysis before a change<\/h3><\/div>
Visibility of data flows across the application<\/h3><\/div>
Onboarding new developers<\/h3><\/div>
Empowering non-technical analysts<\/h3><\/div>
Opening APIs and integrating with modern systems<\/h3><\/div>
The bigger picture: see what DISCOVER reveals about your applications<\/h2><\/div>
![](\"https:\/\/www.arcadsoftware.com\/discover\/wp-content\/uploads\/2026\/06\/monica-yadav.png\")
About the Author<\/h2><\/div>
Monica Yadav<\/h3><\/div>
Solutions Architect, ARCAD Software<\/h4><\/div>