This thesis aims to explore and prototype a system for participatory governance and transparent decision-making in the context of modernizing legacy software systems. Using Large Language Models (LLMs), the system will analyze and extract decisions from various project artifacts (e.g., commit messages, meeting transcripts), represent them in a structured decision domain model, and create an audit trail that supports stakeholder engagement, accountability, and traceability throughout the modernization process. The overall goal is to support more inclusive, explainable, and well-documented modernization strategies by making both decisions and their rationales visible and verifiable to all relevant stakeholders.
Thesis Type |
|
Status |
Open |
Presentation room |
Seminar room I5 6202 |
Supervisor(s) |
Stefan Decker |
Advisor(s) |
Michal Slupczynski |
Contact |
slupczynski@dbis.rwth-aachen.de |
Potential Research Questions
- RQ1: What types of decisions are most important in the context of legacy software modernization?
- RQ2: How can agentic AI systems be used to extract and summarize these decisions from unstructured sources?
What existing knowledge representation formats would be feasible for efficient storage and reasoning? - RQ3: What audit trail format can be used to store and track these decisions?
- RQ4: How useful is such a tool for developers or stakeholders reviewing the modernization process?
Background and Related Work
Modernizing legacy software systems often involves decisions across architecture, code, infrastructure, and compliance—decisions that are typically scattered across unstructured artifacts like emails, issue trackers, and meeting notes. While tools for software modernization and architecture recovery exist, they often neglect how decisions are made, who made them, and why.
Research in software governance, traceability, and AI-supported decision support systems has explored how to model and manage decision-making in software engineering. Recent advances in LLMs now make it feasible to extract and summarize decisions from textual sources automatically. Similarly, audit trails and accountability systems offer mechanisms for immutable tracking and verification.
This thesis combines these ideas to investigate how decisions during software modernization can be captured, structured, tracked, and shared to improve collaboration and transparency.
Expected Contribution
The thesis will deliver a practical and usable prototype that demonstrates how LLMs can be used to support transparent and traceable decision-making in legacy system modernization projects. The results may contribute to broader research in explainable AI, developer collaboration, and tool support for modernization.
- We expect a small-scale prototype that:
- Uses an LLM (e.g., via API) to extract decisions from commits or meeting notes
- Stores decisions in a structured knowledge graph according to some decision model
- Retrieves relevant decision timelines using modern LLM retrieval techniques (e.g. agentic RAG)
- Provides a basic UI to browse/view the decision history
- A decision model tailored to software modernization.
- A short evaluation with example input and user feedback (e.g., developer interview or survey)
If you are interested in this thesis, a related topic or have additional questions, please do not hesitate to send a message to slupczynski@dbis.rwth-aachen.de
Please apply with a meaningful CV and a recent transcript of your academic performance.
Co-Advisor: Ada Slupczynski (SWC RWTH)
- Basic knowledge of software engineering and development workflows (e.g., Git) and software modernization issues
- Basic experience in natural language processing or the use of LLMs, e.g. LangChain or RAGs
- Experience with Python for building simple pipelines, tools or interfaces
- Experience with modern DevOps tools, such as Docker, n8n etc.
- Preferable: Familiarity with JSON, APIs, or front-end basics for UI development