NguyenNhatHao

Software engineering student in Ho Chi Minh City who builds and runs his own scraping services rather than only reading about them.

What I work on

• Web scraping and extraction — Node.js, Python, Crawlee, Docker. I care most about the failures that don't raise errors: an upstream that returns HTTP 200 with half the payload, a directory that serves the organiser's email in place of the exhibitor's, an archived event page that extracts perfectly cleanly and is worthless.
• Production services — tikcatch.com , a video extraction service I built and operate end to end, including the security layer and the incident when a third-party provider quietly stopped returning data.
• Open source — safe-ytdlp , a hardened subprocess wrapper with an SSRF host allowlist, command and argument injection guards, timeouts and concurrency caps. 50 tests, CI across Node 18/20/22.

How I build

Every rule I rely on has a test behind it, and the tests are built from data I captured myself rather than examples I invented. When something is uncertain I measure it and publish the number, including when the number is unflattering.

Find me on GitHub .