Got data n8n can’t easily swallow? Group chat exports, PDFs and scanned docs, websites with no API, spreadsheets nobody dares to touch.
That’s the part I do. I turn it into structured rows (Google Sheets, Postgres, or JSON to your webhook) that drop straight into your n8n workflow.
How I work:
- Python + Playwright for extraction. An LLM only where plain rules can’t do the job.
- Every row gets a confidence flag. Doubtful results land in a review tab instead of being quietly guessed.
- Polite crawling: rate limits, resume after a crash. No captcha-breaking or anything that violates a site’s terms.
First job: fixed $150 for one well-scoped extraction. You watch it run on a sample of your real data before paying. Code and full output are delivered after payment.
Send me where the data lives, a small sample, and what you want to end up with. I’ll reply with a scope and a straight yes/no within 24h.