Lakehouse indexes allow not only to skip files during read but also, thanks to row-level tracking id to go directly to the given row and skip scanning other rows in the file. #databricks
https://www.sunnydata.ai/blog/databricks-full-text-search-index-benchmark
https://medium.com/databrickscommunity/full-text-search-index-on-databricks-i-managed-to-search-400-...

My blog: https://databrickster.medium.com/