Elicit

Elicit operates as an AI-powered research assistant, developed to automate and accelerate complex research workflows for scientists, academics, and other professionals. The company began as a project within the non-profit research organization Ought, co-founded by Andreas Stuhlmüller and Jungwon Byun in 2018, before spinning off as an independent public benefit corporation in September 2023. This transition coincided with a $9 million seed funding round led by Fifty Years, with participation from Basis Set, Illusian, and notable angel investors. A subsequent Series A round in February 2025 raised an additional $22 million, led by Spark Capital and Footwork, valuing the company at $100 million.

The founding team's background is deeply rooted in cognitive science and technology scaling. CEO Andreas Stuhlmüller holds a Ph.D. in Cognitive Science from MIT and was a postdoctoral researcher at Stanford, focusing on machine learning and computational cognitive science. COO Jungwon Byun brings experience in scaling companies, having served as Head of Growth at Upstart, and holds a B.A. in economics from Yale. Their combined expertise drives Elicit's mission to scale up good reasoning by making research more efficient and reliable. The company is structured as a Public Benefit Corporation, underscoring its commitment to prioritizing societal benefit alongside profit.

Elicit's platform serves a client base ranging from individual students and researchers to large organizations like Genentech, Novartis, and various consulting firms. It operates on a freemium business model, offering tiered subscription plans for advanced features. The core service automates time-consuming research tasks by leveraging large language models. Its product allows users to search a database of over 125 million academic papers using natural language queries, going beyond simple keyword matching to find semantically relevant studies. Key features include the ability to extract specific data from papers and synthesize it into organized tables, summarize findings across multiple documents, and upload personal PDFs for analysis. The platform is designed to minimize the risk of AI "hallucinations" by grounding its outputs in verifiable sources, a critical feature for its high-stakes user base in fields like biomedicine and health economics.

Keywords: AI research assistant, literature review automation, scientific research tools, data extraction, academic paper summarization, public benefit corporation, computational cognitive science, evidence-based decision-making, natural language processing for research, systematic review software, research workflow automation, semantic search for papers, AI for science, Jungwon Byun, Andreas Stuhlmüller, research synthesis, academic data analysis, AI-powered knowledge discovery, research productivity tools, machine learning for reasoning