ARTIFICIAL INTELLIGENCE
New Job Board GrepJob Scrapes 67,000 Engineering Listings From Company Career Pages
A Show HN launch promises curated, AI-filtered software engineering jobs pulled hourly from company career sites, with no ghost postings or sponsored listings.
Image: IntraGoals Media · Uploaded by IntraGoals — usage rights confirmed
A new job-search tool called GrepJob has launched on Hacker News, pitching itself as a cleaner alternative to traditional job boards by scraping listings directly from company career pages rather than relying on third-party postings.
According to the project's Show HN submission, titled "GrepJob – 67,000 SWE jobs scraped from a curated list of top companies," the platform claims to pull more than 60,000 tech jobs on an hourly basis from employer career sites. The company says this approach filters out ghost jobs, duplicate reposts, and paid sponsored listings that commonly clutter other job search platforms.
GrepJob uses artificial intelligence to parse job descriptions and extract structured details such as required skills, experience levels, and specific technologies, which the company says powers more accurate search and filtering than keyword-based systems. Example listings surfaced on the platform include engineering productivity roles at Electronic Arts, focused on tooling, automation, build systems, and AI-enabled workflows for game development, and a security leadership position at cloud infrastructure provider Nebius, overseeing detection engineering and incident response across its cloud and bare-metal environments.
In a brief FAQ accompanying the launch, GrepJob's creators said the service focuses on companies with "strong engineering cultures and competitive compensation packages," and that listings are continuously monitored and updated to avoid stale postings.
The claims about job volume, update frequency, and curation methodology come from GrepJob's own promotional materials and have not been independently verified. As with many Show HN launches, the figures and feature descriptions should be treated as the company's self-reported claims pending outside confirmation.