About the data

Every statistic on the public InterviewOS pages — open-role counts, weekly hiring activity, H-1B sponsorship records — comes from the same data engine that powers the product. This page explains where the numbers come from and what they can and cannot tell you.

What the dataset is

Live job listings across roughly 16,000 companies, collected directly from each company's careers site or applicant-tracking-system (ATS) API, combined with public U.S. government records on work-visa sponsorship.

How listings are collected

An automated pipeline refreshes company boards daily, using integrations for about 25 ATS platforms (Greenhouse, Lever, Ashby, Workday, and others) plus browser-based collection for boards that expose no API. Each listing is timestamped the moment the pipeline first sees it, so every company is measured on the same clock.

How listings expire

A listing is marked inactive once it stops appearing on the company's board on re-scrape — typically within about two days of being taken down. A safety guard prevents mass expiry when a scrape looks incomplete, and listings are retained for up to 60 days, so very long-running postings can drop out of the counts.

H-1B sponsorship records

Sponsorship data comes from two public government sources: H-1B petition approvals published in USCIS disclosure data (fiscal years 2021–2023) and Labor Condition Applications certified by the US Department of Labor (current fiscal year), matched to companies in the InterviewOS catalog. Based on public USCIS and Department of Labor records. Past filings do not guarantee future sponsorship.

What the counts mean

Open roles = listings live on the company's board right now, as of the last successful scrape. Added last 30 days = listings first seen in the last 30 days. In the product, momentum = roles added in the last 7 days vs the prior 4-week average. The monthly Hiring Velocity Index compares each company's 30-day additions against its own trailing baseline; its full methodology is published on that page.

Known limitations

• Coverage has gaps: boards with no API and no reliably scrapeable page are missed or refreshed less often, and not every cataloged company has live coverage.
• Posting dates are estimated for boards that don't publish them — the date the pipeline first saw the listing is used instead.
• All counts are point-in-time snapshots; pages are cached for a few hours, so a number can trail the board slightly.
• The catalog is weighted toward technology and venture-backed employers — treat aggregates as directional for this universe, not as national labor statistics.

Using and citing this data

Free to cite with attribution to InterviewOS, linked. Per-company CSV downloads are available on each company page, and the full monthly index dataset at /hiring-index.csv. See Press & data requests.

Contact

Questions, corrections, or data requests: baumn96@gmail.com.