Kensho NERD is an advanced machine learning system designed to extract and link entities—such as companies, people, organizations, events, and places—from unstructured text. By connecting these entities to comprehensive databases like S&P Capital IQ and Wikimedia, NERD transforms raw text into structured, actionable insights, enhancing research capabilities and decision-making processes. Key Features and Functionality: - Entity Recognition and Disambiguation: Accurately identifies entities within text and resolves ambiguities, including abbreviations, acronyms, aliases, and historical names. - Database Integration: Links recognized entities to S&P Capital IQ and Wikimedia databases, providing enriched context and deeper insights. - Context-Aware Processing: Utilizes surrounding text to enhance entity disambiguation, improving accuracy in complex documents. - API Accessibility: Offers a REST API for seamless integration into various workflows, supporting both real-time and batch processing. - Scalability: Capable of processing millions of pages daily, suitable for enterprise-scale operations. Primary Value and Problem Solved: NERD addresses the challenge of deriving meaningful insights from vast amounts of unstructured text data. By automating the identification and linking of entities, it streamlines manual workflows, enhances data discoverability, and empowers organizations to make informed decisions based on comprehensive, structured information.