One-shot extraction
from $500 · 72h
One site, your fields, delivered as CSV, Google Sheets or JSON. Lead lists, product data, reviews, prices. You get the data and the script that produced it, and you own both.
Web data,
mined clean.
The data you need is already sitting on other websites: directories, marketplaces, review pages, price pages. We extract it and hand it over clean, in whatever format you work with. Fixed price from $500. No subscriptions pushed, no meetings about meetings.
Run by Dominique Degottex, the developer behind magpie, an open-source scraping server with proxy rotation, anti-bot browser escalation and 600+ tests.
Built on
Go
TypeScript
PostgreSQL
Node.js
Python
RedisGoTypeScriptPostgreSQLNode.jsPythonRedisWhat we deliver
Four ways to work with us, whether you need a one-off pull, a tricky target, fresh data on a schedule, or the engineering around it. Every job is priced before it starts.
Hard targets
$800–1,500 · 3–5 days
Sites heavy on JavaScript, sites with bot protection, paginations from hell. We drive real browsers, rotate proxies and pace the requests. The stuff that kills DIY scrapers is what we do all day.
Monitoring retainer
$200–500/mo
Data goes stale in weeks. We re-run your extraction on a schedule and deliver fresh rows in the same format. It's optional, cancel anytime, and we never make it a condition.
Custom data engineering
quoted
Go and TypeScript development when you need more than data: APIs, CLIs, pipelines, integrations. Same engineer, same quality bar as magpie.
Why most data projects stall
You ask your team for a quick script. The site changes its layout, and the script quietly breaks. A freelancer quotes three weeks, delivers page one only, then goes quiet. The agency wants a six-week discovery phase before anyone writes a line. The dataset vendor sells you last quarter's numbers in someone else's columns. Three months later, you're still copy-pasting into a spreadsheet. There's a shorter path.
| Criterion | Motherlode | Generalist freelancer | Agency | In-house |
|---|---|---|---|---|
| Start | 72h delivery | 1 to 2 weeks | 2 to 6 weeks | hiring cycle |
| Price | fixed, from $500 | hourly, it drifts | $10k+ minimums | salary |
| Anti-bot work | our core skill | often blocked | subcontracted | varies |
| You own the code | yes, always | sometimes | rarely | yes |
| Scope games | none, fixed quote first | common | change requests | n/a |
From first message to delivery.
Four steps, no black boxes. You see real output early, and you always know where things stand.
Send the target
Just the site URL and the fields you want. A sentence is enough; we handle the recon.
Fixed quote in 24h
Every job starts with a free feasibility check. If we can't get your data cleanly, you hear it from us before you pay a cent.
Sample rows same day
You approve real rows from your site, not a mockup, before we run the full extraction.
Full delivery in 72h
Clean rows in your format, plus the script. A 48h fix window is included.
Proof, not logos
magpie (open source)
A production-grade scraping server written in Go: proxy pool with automatic failover, anti-bot escalation through real browser profiles, screenshots, batch and diff jobs. 600+ tests, all in the open. The same engineering runs your job.
Ask us anything
Every claim on this page can be checked on a 15-minute call. We'll open your target site live and tell you honestly what we can get, what's risky, and what we'd say no to.
Fair questions
The things you should ask before sending us a target, answered straight.
Is web scraping legal?
Extracting publicly available data is legal in most places, and we stay well inside the lines: public pages only, nothing behind logins, no personal data, no busting paywalls or DRM. If a site's terms say no scraping, we tell you plainly during the feasibility check, and we turn down work that crosses the line.
What if the site blocks bots?
That's the job. We rotate proxies, run real browser profiles and keep request rates under the thresholds. It's what magpie was built for.
What exactly do I receive?
Your data in CSV, Google Sheets or JSON, whichever you prefer, plus the extraction script and a short README. Run it again yourself, or hand it to us on a retainer. Both work.
Why not just buy a dataset?
Off-the-shelf datasets are stale and shaped by someone else's schema. Ours is extracted this week, with your exact fields, in your exact format.
What if something's off after delivery?
A 48-hour fix window is included with every job. If a sample was approved and the full run diverges from it, that one's on us. No debate.
Do you sign NDAs?
Yes. Targets, fields and deliverables stay confidential by default, NDA or not.
Ready to see your data?
Send the site and the fields. You'll have a fixed quote within 24 hours. No commitment, no pitch deck.