Transforming GST Compliance: Leveraging AI-Powered Ensemble Fusion Net (EF-Net) for Tamil Nadu's SMEs

The implementation of AI-Powered EnsembleFusionNet (EF-Net) revolutionizes GST compliance for SMEs in Tamil Nadu by automating repetitive processes, enhancing accuracy, and minimizing errors. EF-Net combines Deep Learning, Machine Learning, and Natural Language Processing (NLP) to address key challenges such as tax computation validation, invoice matching, and anomaly detection. The model achieved high accuracy across compliance modules, including 98.7% for tax computation validation, 97.2% for invoice matching, and 95.8% for anomaly detection. These results demonstrate the system's robustness and efficiency in streamlining GST workflows. By automating compliance tasks, EF-Net reduced the average monthly time spent on tax filing by 70%, invoice reconciliation by 73%, and anomaly investigations by 70%. Cost savings ranged from 60-62% across SMEs of varying sizes. EF-Net's hybrid architecture integrates autoencoders for anomaly detection, decision trees for rule-based compliance checks, and NLP for processing unstructured data. This approach ensures comprehensive and scalable compliance solutions tailored to SMEs’ operational needs.