Thoroughly understand, modernize, and validate legacy RPG and COBOL systems.
LIGHTMAN combines deterministic software analysis, step-by-step transformation, and reproducible testing into a modernization platform for traditional development teams and AI-powered workflows.
Deterministic Analysis
Instead of AI Predictions
and without AI
For RPG, COBOL, IBM i, 5250, and complex legacy applications
THE PROBLEM
Legacy modernization rarely fails because of syntax. It fails because of a lack of understanding of the system.
Legacy applications often contain hidden business logic, control flows that are difficult to trace, and complex dependencies between programs, data, and user interfaces. Documentation is either missing or outdated, while critical knowledge is often held by only a few experts.
AI can accelerate modernization projects. However, without a reliable structure, context, and reproducible validation, the necessary foundation for reliable results is lacking.
LIGHTMAN lays this foundation and makes legacy systems transparent, allowing them to be modernized in a step-by-step and controlled manner.
THE LIGHTMAN PHILOSOPHY
First, understand. Then, make changes. After that, prove that it still works.
LIGHTMAN follows a controlled modernization model.
First, structures, control flows, and dependencies are identified. Next, targeted modernization steps are implemented and finally validated using reproducible tests.
Only then are AI agents integrated—with tools, context, and controlled feedback loops.
AI increases the leverage. The technical foundation is built on deterministic analysis and reproducible validation.
THE LIGHTMAN PLATFORM
A platform for the entire modernization journey.
Each LIGHTMAN module solves a specific problem. Together, all the products form a comprehensive modernization path—from the analysis of existing applications through targeted transformations to reproducible validation and the controlled integration of AI.
LEGACY SYSTEMS
RPG / COBOL / OCL / CL / DSPF / DB2 / Zyklus / 5250
UNDERSTAND
MORGANA / PGMWorkflow / Avalon
TRANSFORM
GOTOFree / DSPF2React / RPG2Free / Avalon
VALIDATE
ReplicTest / i and me MCP Server
INTEGRATE AI CONTROLS
i and me MCP Server / AI Production Gate
PRODUCTS
The Orchestrator of Your Modernization
LIGHTMAN serves as the central hub for your modernization efforts. The solution integrates specialized analysis, development, and testing tools and makes their information available to AI agents in a targeted manner. This transforms individual tools into a coordinated and traceable modernization process.
Optimize RPG and COBOL systems through the automatic analysis and secure removal of unused objects.
Features:
- Identifying Unused Programs, Files, and Objects
- Analysis of Dependencies Before Adjustment
- Safe Removal of Unnecessary System Components
Organize large RPG and COBOL codebases directly in VS Code and make them easy to navigate.
Features:
- Outline and Navigation for Legacy Source Code
- Quick access to relevant structural elements
- Support for large and complex programs
Make complex program flows analyzable as annotated control flow graphs.
Features:
- Control Flow Graphs
- Jump and Execution Paths
- Structured artifacts for analysis and documentation
Provide build information, cross-references, and technical relationships as context.
Features:
- Build Context
- Cross-references
- Dependencies Between Artifacts
Make structures and relationships visible.
Converts programs containing jump statements into better-structured control flows.
Features:
- Analysis of Jump Structures
- Transformation of Complex Control Flows
- Creation of Traceable Modernization Artifacts
Modernizes existing 5250 display files into modern React interfaces, including IBM i integration.
Features:
- Analysis of Existing DSPF Surfaces
- Creating React Components
- Backend Integration via OA Handler
Targeted automation of modernization steps.
Transform legacy applications in a controlled and traceable manner.
LIGHTMAN automates selected modernization steps without losing control of the process.
Automated regression testing for 5250, web, and API applications.
Features:
- Tests of 5250, web, API, and other application workflows
- Reproducible test data and comparisons
- Regression testing before and after changes
- Integration into Modernization and Automation Workflows
Stay the course as technology evolves.
Modernization is successful only if existing processes continue to function properly.
LIGHTMAN enables reproducible end-to-end testing for legacy and modernized applications.
Provides analytics and modernization information for AI agents.
Features:
- Standardized use of tools by agents
- Access to Analysis and Modernization Artifacts
- Structured context instead of exclusively unstructured source code
Modernizes existing 5250 display files into modern React interfaces, including IBM i integration.
Features:
- agent-based work cycles
- Integration mit ReplicTest
- Automated Feedback from Tests
- Defined approval and control points
Equip AI agents with tools, context, and validation loops.
AI can accelerate modernization—if it is based on reliable data and reproducible tests.
LIGHTMAN provides the necessary context as well as controlled development and validation processes.
USE CASES
Choose Your Modernization Path
LIGHTMAN can help you get started the right way.
Understanding Legacy Code
01
Related Products:
MORGANA
PGMWorkflow
Avalon
Result:
Quickly identify structures, dependencies, and execution paths.
Applications
02
Related Products:
Avalon
i and me - MCP Server
Blueprint
Ozgar
Result:
Structured technical artifacts as the basis for documentation and knowledge preservation.
Modernizing GOTO-Heavy Programs
03
Related Products:
GOTOFree
ReplicTest
PGMWorkflow
Result:
Analyze and transform complex control flows and validate them against existing behavior.
Modernizing 5250 Interfaces
04
Related Products:
DSPF2React
ReplicTest
Result:
Gradually migrate existing user interfaces to modern web UIs and validate their functionality.
Setting Up Regression Tests
05
Related Products:
ReplicTest
Result:
Ensure that current behavior can be reproduced before beginning the modernization process.
Provide AI agents with legacy context
06
Related Products:
Avalon
i and me - MCP Server
PGMWorkflow
Time-Coverage
Result:
Agents receive structured information and tools instead of just unstructured source code.
Review AI-Generated Changes
07
Related Products:
ReplicTest
AI Production Gate
Result:
Changes are reviewed through defined testing and feedback loops.
Ready for your renovation project?
Talk to us about your specific modernization plan and the relevant LIGHTMAN components.
AI-ready, but not AI-dependent
Ready for AI. Not dependent on AI.
Most LIGHTMAN components already provide direct benefits even without AI. They analyze, structure, transform, or validate software in a deterministic manner.
AI can also utilize and combine these capabilities.
Don't just explain. Show.
LIGHTMAN is a VS Code / Code for IBM i extension that serves as a central modernization platform. It connects all the child tools of the “League of Modernization” to their source code and provides AI coding agents with structured context. As a result, the AI delivers significantly better results while reducing token costs—and replaces mere guesswork with a genuine understanding of your code.
Morgana bridges the gap for IBM i developers who want to switch from RDi (Rational Developer for i) to VS Code using “Code for IBM i.” Morgana includes all of RDi’s outliner functionality and enhances it with practical Features: GOTO markers, mouse-click navigation to procedures in modules/service programs and to programs called via CALL, display of the structures of internally described files, warnings for discrepancies between source code and the object in the LIBL, and tracking of indicators that are modified using MOVEA.
PGM Workflow enables complex program flows on IBM i to be analyzed as annotated control flow graphs. It visualizes and structures the actual execution paths (including jumps and execution paths) and uses them to generate clear, documentable artifacts. The operating model is AI-optional—the core functions work even without artificial intelligence.
Avalon is a build and deployment toolchain whose primary function is to migrate source code files from the traditional QSYS (Source Members) to the IFS. This makes the source code Git-compatible and prepares the system for modernization.
In addition, Avalon provides build information, cross-references, and technical relationships as context, highlights dependencies between artifacts, and makes cross-references visible so that AI systems can better understand the technical relationships between programs and objects.
DSPF2 → React - AI Tool LIGHTMAN
→
Old 5250 Screen
New React UI
DSPF2React is an AI-powered module from i and me that automatically converts entire business processes—such as those consisting of one or more 5250 display files (green screens)—into modern React components.
The existing RPG or COBOL code remains unchanged. The result is a modern, browser-based user interface that can be expanded with web services and custom features as needed, while still remaining connected to the original application.
GOTO-Free converts programs containing jump statements (GOTO) into better-structured control flows. It analyzes existing jump structures, transforms complex control flows, and generates traceable modernization artifacts in the process. The tool operates autonomously and supports the cleanup and modernization of legacy code on IBM i.
ReplicTest is an automated testing tool for IBM i applications.
It enables testing of native 5250 and batch applications, web services, and modern web interfaces on a single, integrated platform. Tests can be recorded interactively or executed as pre-built scripts—reliably, repeatably, and without having to change existing development workflows. In addition, it offers RPGUNIT integration, code and time coverage, and deterministic test data.
AI Production Gate: The Missing Control Layer in the AI Stack.
It acts as an independent compliance authority that uses ReplicTest to deterministically verify AI-generated code transformations against the original behavior and, if necessary, returns errors and input to the AI agent. In addition, it approves or blocks deployments, generates auditable evidence of functional integrity, and thus enables the secure, traceable transfer of AI changes to production.
Get in touch.
Connecting Through Knowledge.
Digital IT training, learning platforms, and customized solutions to help you advance.