A domain-agnostic, LLM-based pipeline developed to automate the screening phase of high-volume Systematic Literature Reviews. The tool implements a dual-agent architecture (Assistant and Evaluator) with structured conflict resolution and inclusion-biased defaults to minimize false negatives. It supports bibliographic data from ACM, IEEE, Scopus, and Springer, and handles the full workflow from raw data normalization and abstract augmentation to multi-phase agentic screening and relevance selection. Developed and validated in the context of a Systematic Literature Review on Agentic AI in Software Product Development.
Paper
The full text of this publication is not hosted on 44B due to licensing.
Read it at OpenAlex