
I took a quick look and with the way they’re published, that’s going to be a bit difficult. You’ll probably have to Ctrl-F through each batch. What sucks about that is that your browser will have to look through all the text by itself, which some browsers don’t do super well (or it’s slow on some computers). With the Epstein files, the DOJ indexed the files and took you right to the one containing that text (as required by the transparency act)
If you’re doing this for yourself and want to ask an LLM questions, you could make something RAG-based with something like AnythingLLM (haven’t tried it though). If you want something like Jmail or the DOJ’s site, you’ll need to extract the text from the documents and put it in something like Elasticsearch (can be a bit difficult sometimes) or Typesense (easier)