TigZig serves interactive dashboards over hundreds of millions of rows by using DuckDB as an embedded analytical engine behind a FastAPI backend. DuckDB runs the heavy aggregations in-process (no separate database server), and FastAPI serves the results to the frontend, which keeps queries fast even on large tables. A live example is the Movie Explorer dashboard built on IMDb data (12M+ titles, 23M+ rows, 16GB).
The open-source backend is at https://github.com/amararun/shared-duckdb-dashboards-backend. A write-up of the approach (a custom dashboard on 230M+ rows with DuckDB and FastAPI) is at https://www.tigzig.com/post/custom-dashboard-duckdb-fastapi-230-million-rows.
Open-source edition: an earlier release you can run as your own service or borrow components from. Treat it as a reference implementation rather than a mirror of what is running here - the hosted service has moved on since.
Building something like this? How I work covers the rates, the availability and what I take on.