Show HN: Use Claude Code to Query 600 GB Indexes over Hacker News, ArXiv, etc. (exopriors.com)
There's also an Alerts functionality, where you can just ask Claude to submit a SQL query as an alert, and you'll be emailed when the ultra nuanced criteria is met (and the output changes). Like I want to know when somebody posts about "estrogen" in a psychoactive context, or enough biology metaphors when talking about building infrastructure.
Currently have embedded: posts: 1.4M / 4.6M comments: 15.6M / 38M That's with Voyage-3.5-lite. And you can do amazing compositional vector search, like search @FTX_crisis - (@guilt_tone - @guilt_topic) to find writing that was about the FTX crisis and distinctly without guilty tones, but that can mention "guilt".
I can embed everything and all the other sources for cheap, I just literally don't have the money.
40 comments
[ 1.9 ms ] story [ 64.5 ms ] thread[1] https://github.com/giatenica/gia-agentic-short
Hopefully your API doesn't get exploited and you are doing timeouts/sandboxing -- it'd be easy to do a massive join on this.
I also have a question mostly stemming from me being not knowledgeable in the area -- have you noticed any semantic bleeding when research is done between your datasets? e.g., "optimization" probably means different things under ArXiv, LessWrong, and HN. Wondering if vector searches account for this given a more specific question.
Okaaaaaaay....
what makes this state of the art?
1) Full readonly-SQL + vector manipulation in a live public database. Most vector DB products expose a much narrower search API. Basically only a few enterprise level services let you run arbitrary SQL on remote machines. Google BigQuery gives users SQL power, but it mostly doesn't have embeddings, connect public corpora, have as good of indexes, and doesn't have support an agentic research experience. Beyond object-level research, Scry a good tool for exploring and acquiring intuitions about embedding-space.
2) An agent-native text-to-SQL + lexical + semantic deep research workflow. We have a prompt that's been heavily optimized for taking full advantage of our machine and Claude Code for exploration and answering nuanced questions. Claude fires off many exploratory queries and builds towards really big queries that lean on the SQL query planner. You can interrupt at any time. You have the compute limits to do lots of exhaustive exploration--often more epistemically powerful than finding a document often, is being confident than one doesn't exist.
3) dozens of public commons in one database, with embeddings.
How much do you need for the various leaks, like the paradise papers, the panama papers, the offshore leajay, the Bahamas leaks, the fincen files, the Uber files, etc. and what's your Venmo?
As it is, the Supplementary Materials are inconsistently indexed so a lot of insight you might get from the last 15 years of genomics or proteomics work is invisible.
I imagine this approach could work, especially for Open Access data?
Using LLm for tasks that could be done faster with traditional algorithmic approaches seems wasteful, but this is one of the few legitimate cases where embeddings are doing something classical IR literally cannot. You could also make make the LLM explain the query it’s about to run. Before execution:
“Here’s the SQL and semantic filters I’m about to apply. Does this match your intent?”