Supplementary Data for SLR: Agentic AI in Credit Risk Management - A Predictive Framework for Early Detection of Non-Performing Loans in Microfinance Institutions

This repository contains the comprehensive supplementary dataset for a Systematic Literature Review (SLR) focusing on the integration of Agentic AI inside Credit Risk Management, specifically for the early detection of Non-Performing Loans (NPLs) within Microfinance Institutions (MFIs). The dataset aggregates and synthesizes metadata from 200 scholarly articles extracted via semantic search (2021–2026) and screens them down to 80 included studies following the PRISMA 2020 guidelines. The spreadsheet structure is organized as follows: 1. Cover & Dashboard: High-level overview and key metrics of the SLR study. 2. PRISMA Flow Diagram: Quantitative mapping of the selection process (200 identified -> 120 excluded -> 80 qualified). 3. Search Strategy Matrix: Database queries used across major platforms. 4. RQ Framework: The mapping of 5 critical Research Questions (from technological trends to non-technical adoption criteria). 5. Literature Stream: Analysis of 4 distinct streams (S1: ML Credit, S2: XAI Credit, S3: Agentic AI, S4: Excluded Non-Financial). 6. Data Extraction & Literature Review: Full detailed technical metadata including dataset sizes, AI techniques used, and specific performance metrics. This dataset confirms a critical research gap: 0 articles currently address the convergence of Agentic AI, NPL prediction, and Microfinance frameworks simultaneously. Prepared for ICTIM 2026 Track 5.