Replication Package - Applying RAG in Industry: Insights from Grey Literature and Practitioner Perspectives

Replication Package This replication package accompanies our interview study investigating the adoption of Retrieval-Augmented Generation (RAG) in industrial settings. It contains all study materials, extracted data, analysis artifacts, and scripts required to reproduce the results presented in the paper. The file Use_Case_table.pdf contains the complete list of all identified use cases, including their descriptions, assigned classifications, and relevant contextual annotations used during the analysis. We structured the replication package into two main parts to clearly separate newly generated materials from artifacts originating from our prior work. Part 1: Grey Literature Review (GLR) Materials (New Contributions)This part contains all newly created artifacts related to the Grey Literature Review. It includes: The complete set of identified and analyzed sources A detailed description of the GLR process (search strategy, selection criteria, and analysis procedure) Extraction sheets and classification files These materials fully document the GLR workflow and enable transparent reproduction of the review and classification process. Part 2: Artifacts from Our Previous StudyThe second part contains files that originate from our previously published paper:https://www.scitepress.org/publishedPapers/2025/137395/pdf/index.html These files include the original interview materials, research question–specific findings, and supporting analysis artifacts from the earlier study. They are provided here to ensure completeness and reproducibility, and to clearly document the foundation upon which the extended GLR-based analysis builds. Overview of Contents (GLR) File Name Description glr.pdf Description of the Grey Literature Review (GLR) process glr-extraction.pdf Extracted data and coding results from the GLR use_case_extraction_table.xlsx Supporting extraction and coding spreadsheet used during analysis (GLR) extracted_challenges.csv LLM-extracted challenges with descriptions and assigned categories extraxtion_goals.csv Extracted items that could not be confidently categorized benefit_category_counts.csv Aggregated counts of categorized benefits challenge_extraction.py Script used to extract and categorize reported challenges from interview transcripts goals_extraction.py Script used to extract and categorize use cases and goals from interview transcripts Overview of Contents (Interviews) File Name Description RQ1_Usecases.pdf Findings related to Research Question 1: Industry use cases and goals RQ3_Goals.pdf Findings related to Research Question 3: Goals RQ3_Challanges_LessonsLearned.pdf Findings related to Research Question 3: Challenges and lessons learned RQ4_requiremnts.pdf Findings related to Research Question 4: Requirements RQ4_Evaluation.pdf Findings related to Research Question 4: Evaluation practices RQ4_rapid_literature_search.xlsx Extraction table from the rapid literature review supporting RQ4 Interview Analysis Procedure.pdf Detailed description of the qualitative interview analysis process interview_questions_englisch.pdf Interview guide used in English interview_questions_german.pdf Interview guide used in German RQ4_rapid_literature_search.xlsx Structured extraction table from the rapid literature search