Give your AI agent
a live, versioned view of any website.
Describe the fields you want, in plain English. Quorel extracts them nightly, keeps every past version queryable, and hands your agent an MCP tool that can read and fix it directly, without you writing a scraper.
No account needed to read a public dataset. Upgrade only when you're publishing your own.
> what changed on job-board-listings this week
→ queried v14 vs v9, 23 postings added, 8 removed
> fix the missing salary fields and publish v15
→ reviewed 200 rows, filled 14 fields from listing text
→ live at /job-board-listings/v15
LIVE, NOT A MOCKUP
This is a real public dataset, right now
Pulled live from a Quorel dataset of remote job listings. Nothing staged, nothing pre-cleaned. Query it directly, or watch an agent work on it.
Public dataset, no login needed. Browse more datasets
FOR AI AGENTS
An agent can query and clean your data, not just read it
Every dataset ships with an MCP server, on every plan. Claude, or any MCP-compatible agent, can filter it, sort it, and even fix bad rows, without you writing a script to do it.
Ask it questions directly
"What changed in this dataset since last week" just works. No query language to learn.
Let it fix what's broken
An agent can pull the raw rows, correct nulls or duplicates, and publish the result as a new version.
Same safety net either way
Agent-made changes land in the same versioned, reversible system as anything you'd do by hand.
> find listings with price drops this week
→ 47 rows matched, sorted by change
> fix the null prices and publish v3
→ reviewed 1,204 rows, fixed 12 nulls
→ live at /listings/v3
Included on every plan, including Free.
WHY IT'S DIFFERENT
Most data tools stop at the present.
We hand you the past too.
They hand you today's snapshot and move on. Quorel keeps every past version, permanently queryable, so you can see what changed and when, not just what's true right now.
BEFORE
Write a scraper for one page layout
It breaks when the site changes the markup
Write cleanup scripts for nulls and duplicates
No record of what the data looked like last week
AFTER
Describe the fields you want, in English
Extraction adapts when the page layout changes
Every value arrives typed and deduplicated
Every nightly refresh is a saved, diffable version
See exactly what changed
Compare any two versions side by side: added rows, removed rows, changed values.
Roll back in one click
A bad refresh pulled in dirty data? Revert instantly, or freeze a version so it can't change again.
Fork any public dataset
Start from someone else's dataset the way you'd fork a repo. Add your own fields, publish your own copy.
USE CASES
What people actually build with it
Competitor pricing history
See every price change, not just today's number.
Job market monitoring
Track postings, salaries, and what got taken down.
Real estate over time
Status changes, price drops, days on market, versioned.
Live context for your agent
Let Claude check what changed on a site since last week.
A data backend for your app
Query params in, structured API out. No scraping infra to own.
PRICING
Flat pricing, no usage bills
Start free. Upgrade when one dataset stops being enough.
Free
For trying it out.
Pro
For running data pipelines you rely on.
Scale
For teams running this at real volume.
INCLUDED ON EVERY PLAN
FAQ
Common questions
Your first dataset takes
about five minutes.
No credit card, no infrastructure to set up. Or start from an existing public dataset and fork it to fit what you need.
