Diogo O.

Most scrapers work on day one. Mine are built for day 300. I build and maintain Python data systems for sites that fight back: bot protection, JavaScript rendering, rate limits, logins and layouts that change without warning. My production scrapers checkpoint, resume after failures and alert me when a source changes, so fixes land before you notice missing data. What I deliver: → Data from protected and JavaScript-heavy sites: Playwright, session and fingerprint handling, proxy rotation, or reverse-engineering the site's internal API so no browser is needed → Scheduled pipelines delivering clean CSV, Excel, JSON or straight into your database → Bulk public-records and directory extractions, reconciled against the source's own totals → Document and OCR pipelines (1,099 registry documents processed in one project) → The backend around the data: FastAPI services, PostgreSQL, Docker deployments on your server → LLM features on your data: classification, summaries, document Q&A with cited sources How I work: fixed price per milestone, a small paid pilot first on larger jobs, and an optional monthly upkeep plan so the data keeps arriving when sites change. Send me the site and the fields you need, and I'll tell you what will be hard about it before you pay anything.