Back to Projects 
October 2025
PECV-bench
Research Benchmark
Benchmark for LLM-based detection of cross-artifact inconsistencies in programming exercises. 91 variants, 93 labeled issues, multi-model evaluation.
Technologies
PythonLangChainLLMJava
Related Projects
EduTelligence (Athena)
Apr 2023Athena is the automated assessment component of EduTelligence, a suite of AI services for the Artemis learning platform. I designed the modular architecture and built feedback generation for text exercises.
PythonFastAPILLMLangChainDocker +1
LLMs for Automated Feedback Generation
Apr 2023Integrating LLMs for feedback generation into Artemis and setting up a research playground for quick experimentation and iteration.
PythonJavaTypeScriptNLPLLM +1