Development of a Large Language Model (LLM) for Project Tagging

The Education Global Practice of the World Bank seeks to develop a Large Language Model (LLM) capable of analyzing project documentation and automatically tagging projects and analytical pieces (report, papers, briefs, etc.) based on a predefined taxonomy of activities and topics. The taxonomy will be provided by the WB team. The goal is to enhance the categorization, retrieval, and analysis of education project and analytical work data. A small firm with expertise in natural language processing (NLP), machine learning, and AI model development is sought to build, test, and deploy this system.