Keenable SELECT: an agent that searches the web in SQL
keenableai.github.io"One user, inigyou, out-commented all of Hacker News in August 2026 — 3,344 comments, 4.4× the runner-up" - https://select.keenable.ai/artifacts/MarIEMVi6t63M4qovQ2cVw Prompt was "What user on Hacker News made most comments in August, 2026?". Seemed pretty fast, used reasonable sources, results makes sense, pretty good job done by the people building it :)
Some others:
- "What website/domain was the most popular on the frontpage of Hacker News ..." - https://select.keenable.ai/artifacts/SITl-y08FZxJLnif_c_fSw (seems reasonable at a glance)
- "How has the sentiment of PG shifted on Hacker News between 2010 and 2026?" - https://select.keenable.ai/artifacts/pmW5bxWt4t1CgniaF8kBeg (less good, found too little data, moved ahead with an analysis anyways)
I wish there was a way to abort/cancel/edit the requests, as I am, like many of you, still human, and therefore do mistakes sometimes, but I do realize them, although sometimes a bit too late.
Thanks for the feedback on the cancel/abort/edit requests. we are working on this already.
inigyou claims (1) to be a sockpuppet of a previously banned user (2) to be now banned/restricted and (3) to have a username based on a slur, which I have not been able to figure out, despite diligent LLM assistance.
Assuming this is accurate, the diagram showing movement between different AI labs is pretty interesting: https://select.keenable.ai/r/frontier-ai-researcher-moves-si...
Suggests that when Meta AI started throwing around those multi-million-dollar paychecks they got 12 from OpenAI, 5 from DeepMind, 2 from Thinking Machines Lab and none from Anthropic.
(There's a small print disclaimer about accuracy at the bottom: "This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors.")
Hey, I'm on the team working on this. In terms of accuracy the best way to gauge the usefulness of these reports is to search for some you are deeply familiar with and then estimate.
In general I find these reports have some gaps but for most topics they provide a reasonable picture of what is going on. I personally prefer them over deep research reports coming out of chatgpt etc. as they are just walls of text that I skip to go to the result table.
For example as somebody interested in coffee this is pretty comprehensive based on my previous research work: https://select.keenable.ai/r/dual-boiler-prosumer-espresso-m...
Love seeing people experiment with stuff like YQL (https://en.wikipedia.org/wiki/Yahoo_Query_Language) taking advantage of all the AI stuff definitely increases the power.
Echoes of a semantic Web that never was! Showing the prompt trajectories is also a nice touch.
Hey! I’m on the team working on this. The semantic web comparison is pretty close to how we think about this.
One key difference is that we’re using LLMs to create structured data on the fly based on the query. That means we don't need to relying on webmasters to explicitly annotate pages with metadata which in my view was one of the main reasons why the semantic web did not work out.
SEM_EXTRACT, SEM_MATCH, WEB_SEARCH are pretty vague functions. Makes me wonder why you'd bother with the SQL part at that point and not just ask the question.
You need ways to sift through large amounts of unstructured data and do data analysis after. We created these functions to be able to use LLMs to parse unstructured data from the into structured data which we then process with a SQL like engine to map/group/filter etc.
Looks like it thinks rather highly of itself - https://select.keenable.ai/r/the-agent-native-rebuild-of-the....
Otherwise looks really cool!
SQL shall never die, in contrast, it shall grow in scope and utility