Much of pharma’s data is stored in large, complex, and siloed data sets. 80% of that data is unstructured and must be manually curated. Manual data curation, however, is extremely time-consuming and error prone, which slows the drug development process, leaving many research questions unanswered. Our natural language processing (NLP) platform replaces and automates manual curation of pharma’s disparate data sets. It provides a user-friendly, Google-like question and answer experience. Our unique platform connects and searches multiple, disparate, unstructured data sources, returning answers to questions in seconds. This “Expert-In-The-Loop” machine learning approach is optimized to help turn pharma domain experts into data scientists. It is an ideal tool for RWE generation for use in clinical development, medical affairs, post-market surveillance, and HEOR. An HEOR literature review recently found that our platform yielded 10 times the results of a manual search and shortened the review time from weeks to minutes. Our platform has also been validated for accuracy by several global pharma companies who have used it to obtain RWE to map the patient journey and HCP activity, yielding significantly more focused data in a fraction of the time.