Scientific publications are among the most valuable achievements of mankind, but existing methods for navigating and assimilating the rapidly growing literature are still limited and are often closed source. The aim of this project is to assist scientists with discovering, characterizing, and assimilating the knowledge stored in open access (OA) publications and also to streamline the process of scientific discovery and writing. To this end, we have developed a web-based application ("Endoc") with tools for finding the most semantically relevant papers, summarizing long texts, generating original discussions, and more -- all of which are powered by state-of-the-art Natural Language Processing (NLP) techniques. Users can work with very recently published papers. To support Open Research Data (ORD) practices, we have released the code for processing papers, building databases, and NLP tools through a publication at ACL (Gu and Hahnloser, 2022), and we enabled developer access to our services via an application programming interface (API). In addition, we also established the Endocite Association, which makes it possible for the research community outside of UZH, ETHZ, and EPFL to benefit from Endoc as well.