Volodymyr A.

I build web scrapers and data-extraction pipelines that keep working: websites, PDFs, invoices and scanned documents turned into clean, structured data in your spreadsheet, database or CRM. If someone on your team copies data off websites every morning, retypes invoices into a spreadsheet, or runs the same report by hand every week, that's the work I take off your plate. WHAT I DO • Web scraping & data pipelines Python scrapers (Scrapy, Playwright, Selenium) for static and JavaScript-heavy sites, run once or on a schedule. Cleaning, de-duplication and delivery to Google Sheets, Excel/CSV, Airtable, PostgreSQL or your API. Logging and alerts, so you hear about it when a site changes instead of finding gaps in your data later. • PDF, invoice & document extraction Invoices, purchase orders, statements and scanned PDFs turned into structured CSV, Excel or JSON, using OCR and LLM-based parsing with validation rules, plus a simple review screen for anything the system isn't sure about. • Fixing and finishing existing systems Debugging broken apps, scripts and servers, migrations, and getting AI-built apps (for example, made with Lovable) ready for production. I'm used to being called in when something is down. • Automations around your data n8n, Make and Python integrations, and AI agents that classify, route and act on the data once it's extracted. RECENT WORK ON UPWORK • 150 hours so far on an hourly web-scraping project in Python and Java. • Invoice data-entry tool: each PDF shown next to a structured form, so invoices are keyed in one place. "Delivered the project exactly as required and was very responsive throughout the entire process." • Large-batch PDF processing: automatic header/footer removal across a big set of documents. "Perfect job, I recommend him strongly." • Overnight rescue of a corrupted Windows server on Azure while the client's own customer was out of business. "Volodymyr was a virtual superman... He worked throughout the night." • Improvements to a ZXING-JS barcode-scanner web app. HOW I WORK 1. You send the site or a few sample documents and the fields you need. 2. I reply with a sample of the output and a fixed price or an hourly estimate. 3. You get working code with error handling, a short handover doc, and optional maintenance. I work within websites' terms and data-protection rules: public business data only, no personal-data harvesting or spam tools. Tell me what you're still copying by hand, and I'll tell you straight whether I can automate it and roughly how long it'll take.