We propose THELMA (Task Based Holistic Evaluation of Large Language Model Applications), a reference free framework for RAG (Retrieval Augmented generation) based question answering (QA) applications. THELMA consist of six interdependent metrics specifically designed for holistic, fine grained evaluation of RAG QA applications. THELMA framework helps developers and application owners evaluate, monitor and improve end to end RAG QA pipelines without requiring labelled sources or referencethis http URLalso present our findings on the interplay of the proposed THELMA metrics, which can be interpreted to identify the specific RAG component needing improvement in QA applications.
View on arXiv@article{patel2025_2505.11626,
title={ THELMA: Task Based Holistic Evaluation of Large Language Model Applications-RAG Question Answering },
author={ Udita Patel and Rutu Mulkar and Jay Roberts and Cibi Chakravarthy Senthilkumar and Sujay Gandhi and Xiaofei Zheng and Naumaan Nayyar and Rafael Castrillo },
journal={arXiv preprint arXiv:2505.11626},
year={ 2025 }
}