Software Engineer · Jyaba
Dec 2025 — nowPokhara, NepalI work on the data side: getting information off the web reliably and into a shape other systems — and AI agents — can use.
- Write production Scrapy spiders day to day: 110+ so far, across retail, reviews, jobs, real estate, finance and research sites. API-first, with a browser only when needed.
- Built an agent pipeline that turns a URL into a working Scrapy spider.
- Wrote MCP servers and extended Chrome DevTools MCP for AI agents.
- Built a layered data pipeline on S3 and PostgreSQL, with LLM extraction.
- Rebuilt contact extraction for a legal-PDF service, with OCR for scanned pages.
- Build and package a patched Chromium for Scrapy and Playwright.