Hotels & hospitality
Track competitor rates on Booking, Airbnb, and direct sites. Get pinged when prices move, not on every run. Mid-week board meeting brief writes itself.
Cast your net on the web.
Watch any URL. AI alerts only when something actually changes.
For businesses
Trawl Watch. Wake up to signal.
For business teams who need signal, not scripts. Trawl watches the pages that matter. AI judges before alerting.
Watch
King Suite
189169 EUR-11%
4 rooms left
Paste any URL: a hotel listing, a competitor's pricing page, a job board. Describe what you want to track in plain English. No selectors, no scripts.
Every morning, a clean digest lands in your inbox: prices that dropped, stock that came back, new listings that match your filters. A 7-day sparkline per scrap shows how the signal trends.
AI judges what moved, what reversed, and what's worth acting on. The morning we stay silent is the morning nothing actually changed.
For developers and agents
Trawl Dispatch. Your scraper, as an API.
Every scraper exposes a live endpoint. Schedule it, route it through smart proxy tiers, ship results to a webhook, or drive it from your terminal. Built for developers and agents alike.
~/scraps›trawl fix hotels.booking
↳ reading selector…
-price = $('.old-price')
+price = $('[data-price]')
✓ patched · scrap re-ran in 2.4s
GET/api/scraps/hotels.booking200
{ "price": 142, "stock": "in", "last_run": "07:02:15Z" }
Smart proxy routing picks the cheapest legitimate path per run and escalates only when needed. Anti-bot resolution and region-aware IPs included. Schedule at any cadence.
AI auto-fix ships selector patches from a shared cross-tenant pattern library. Most fixes land before you notice. Failed runs retry automatically; the library learns from every patch.
Drive it from any agent or any pipe. The MCP-ready Skill diffs selector changes from your terminal. The CLI runs scraps in any CI job. Webhooks push results into your pipeline. Natural language in, patched code out.
Use cases
A few examples. Every URL is one more.
Where teams ship Trawl today, across business and developer tooling. AI judges before alerting. You only hear from us when something actually moved.
Ground Claude in your data
Live scraper endpoints any MCP agent can query. Drop stale context for fresh pages. The Skill installs in one command.
Business
E-commerce & marketplace
A competitor SKU back in stock, a price drop, a new variant. You hear about it once, the moment it happens.
Business
Jobs & procurement
Tender boards, job postings, supplier pages. Never miss a window where hours matter.
Business
Travel & mobility
Flights, train slots, car rentals on the routes you watch. Alerted only when a deal lands.
Devs & agents
Know your ICP
Prospects, company sites, socials. Fresh B2B signal on demand, fed straight into your growth pipeline.
Devs & agents
Smoke-test critical flows
Point Trawl at checkout, login, pricing. Get paged the minute the real HTML stops matching.
Devs & agents
Run scraps from CI/CD
One CLI install, one trawl dispatch line in your job. Push fresh signal into any pipeline without standing up scrapers.
How it works
Three steps. Zero code.
Paste, sleep, skim. The whole loop fits on a napkin, and scales to thousands of runs.
STEP 01 · PASTE
URLtradewind-resort.example/rooms
DESCRIBE WHAT TO WATCH Watch nightly rate for a king room in October.
Start watch
01
Paste any URL
Drop a hotel listing, product page, or job board into the form. Describe what to watch in plain English. No selectors, no scripts.
STEP 02 · WATCH
09:00tradewind-resort.exampleT0ok
10:00altair-airlines.exampleT1ok
11:00hexa-careers.exampleT3running
stealth · every hour · auto-retry
02
We watch while you sleep
Every hour, every day. Stealth tiers ramp automatically to keep runs clean, and CAPTCHA is solved for you. You stay logged out and asleep.
digest.trawl.me07:12 AM
3 changes overnight
- −16% Hotel rate dropped €489 → €412
- +3 New listings match your filters
- in Gaming bundle back in stock
03
Wake up to a clean digest
A bullet-pointed email of what actually changed overnight: price drops, new listings, stock back in. Skim in 30 seconds.
0%
AI fix accuracy
<0s
Average auto-fix
0/7
Scheduled monitoring
Last run
Only data we keep
AI & Agent suite
AI workflows. One platform.
Where agents drive Trawl end-to-end. Spin up, fix, follow, learn. From one chat, one CLI, one shared brain.
$ claude >scrap rooms on tradewind-resort.example
↳ drafting selectors…
↳ running test…
✓ trawl run shipped in 47s
Spin up scrapers from your agent.
Install @trawlme/cli plus @trawlme/skills. Ask Claude or Codex to scrape a URL, the agent drafts, tests locally through the skill, and ships the scrap directly into Trawl.
$ codex >why is this scrap empty?
- .price-old { selector: '.price' }
+ .price-old { selector: '.price-was' }
✓ patched, re-ran clean · 2.4s
Ask your AI to fix it.
When a scrap breaks, Trawl's UI agent patches from the shared knowledge base. Your CLI agent gets the precise error snapshot and patches the scrap on its side. No tickets either way.
$claude > last 5 runs tradewind-resort
✓07:42812ms· 42 rooms · €169 min
✓06:42790ms· 42 rooms · €169 min
⚠05:423.2s· empty result, auto-fixed
✓04:42812ms· 41 rooms · €172 min
✓03:42830ms· 41 rooms · €172 min
Follow history in chat.
Ask your agent for the last runs and Trawl streams status, durations, and previews back into the conversation. The run journal lives where you already work.
shared KB
selector.price
selector.stock
pattern.list
pattern.detail
fix.timeout
fix.retry
org_xyz
org_abc
Shared FIX + scrap knowledge.
Every patch and every selector signature feeds a cross-tenant knowledge base. The next site that breaks the same way gets fixed faster, for everyone.
Inside Trawl
One scrap. End-to-end.
What runs under the hood when you point Trawl at a URL. Explained for business teams, with a docs link per stage when you want to go deeper.
01
Onboard in under ten minutes.
The AI Wizard reads any URL and emits a runnable headless-browser script in seconds. Refine it inline with the built-in editor and name every scrap so your dashboard stays organised. Persist credentials per scrap in the encrypted credential vault. Authenticated targets keep working without re-logging in. The multi-step onboarding wizard walks you through import, authoring, credentials, and your first run.
Read about Trawl securitywizard · onboarding
✓Paste a URL, describe what to watch
✓AI Wizard generates a runnable script
▶Name your scrap · first run in <10 min
✓ onboarding complete · first scrap running
FAQ
Questions & answers
Everything about plans, smart routing, compliance, and how Trawl behaves when a site changes. If your question is not here, ping us. We update this list every week.
Ready to start
Cast your net.
Start free. No credit card. No code required. Paste a URL, describe what matters, wake up to a clean digest tomorrow.