Certain terms or searches could be associated with a URL pattern. Essentially an automated "blog" or sorts. These could then have schema.org feed types and properties used in it.
It is interesting to see how this compares to something like the Streamlit/Dash dashboard concept we are exploring in other locations.