GXL, also known as Generative Expert Labs, is an applied research lab focused on making scientific knowledge accessible and useful to AI agents. Cofounded by Stanford Professor James Zou and Christine Lemke, the organization works to equip agents to support scientific discovery and advance biomedicine. Its approach addresses a core challenge: much of the world’s scientific information was not designed for AI systems. GXL’s tools make this information native to AI, with the aim of helping agents work more reliably while improving performance, cost, and transparency.

Its first product, Paperclip, provides access to scientific and biomedical resources, including millions of full-text preprints, regulatory documents, clinical trials, patents, and biological databases such as UniProt, PDB, and ChEMBL. Paperclip CLI lets agents explore and synthesize this material through familiar filesystem commands. Paperclip Chat offers a conversational agent that can find resources, organize them in a workspace, and work with users’ uploaded files and figures in research workflows. GXL also shares updates and insights through its blog.