The resources required for access, storage, and preservation of research data can be overwhelming at small and mid-sized institutions. This lightning talk reports on a suggested alternative. With funding from Institute of Museum and Library Services (IMLS), our team built a prototype for a lightweight, open source Dataset Search that promotes discovery of research datasets that are hosted by third-party data repositories. Our Dataset Search complements Google Dataset Search, SHARE, DataMed, NYU Data Catalog, and other research data indexes, adding to the conversation a three-pronged focus. First, our tool promotes discovery for institution-specific research datasets, and includes analytics dashboards that allow institutions to showcase research data as a scholarly product and a driver of institutional reputation. Second, Dataset Search provides enhanced descriptive metadata for individual datasets, produced through topic mining of scholarly profiles like ORCID. Third, Dataset Search metadata is optimized for discovery by commercial search engines. The Phase 1 prototype code will be published on Github in December 2019, including automated setup and instructions for adjusting the tool for local implementation. For Phase 2, starting in January 2020, we will join the Data Discovery Collaboration Project (formerly Data Catalog Collaboration Project) to support community partnerships. We will also partner with a local research center focused on Indigenous and rural health to pilot a feature that will allow the center’s partners to discover restricted datasets relating to their communities. Dataset Search is a contribution to community-driven, community-owned infrastructure for discovery of academic institutional research data.