Improve modern AI-based testing frameworks by adding a formal layer in the background via model learning.
Automata Learning for Coverage, Heuristics, and Evaluation of Models (ALCHEM)
Partner Call open until: May 8, 2026
Project Start: May 2026
Apply model learning to build a formal model during the LLM based testing, to ensure
- Sufficient coverage for regulatory compliance
- Minimizing manual validation
- Explainability of the LLM decisions
Expected results
- Model reflecting an abstraction of the existing test cases
- Coverage analysis based on the model
- Flagging of unusual test results
