Скачать или смотреть AI/ML-based Automated Evaluation of R&D Proposals

Our idea focuses on developing an AI/ML-based Auto-Evaluation System designed to streamline and accelerate the assessment process for R&D proposals submitted to NaCCER, CMPDI Ranchi. At present, the evaluation of proposals is a completely manual process—slow, labor-intensive, and often inconsistent due to subjective judgment. To address these challenges, our system introduces automation through advanced Natural Language Processing (NLP) and Machine Learning (ML) models such as BERT and SBERT, which enable semantic understanding and comparison of proposal content. The system leverages FAISS for fast and efficient information retrieval, allowing new R&D submissions to be quickly compared with previously approved proposals and research datasets. This helps in assessing the novelty, relevance, and compliance of each submission with existing standards and guidelines. In addition, it includes an automated financial verification module and an analytics dashboard that provides transparent, data-driven insights to reviewers. By integrating these technologies, the system ensures a faster, more consistent, and objective evaluation process. It not only enhances efficiency and accuracy but also promotes transparency and accountability through audit trails and automated reporting