Package registries
npm registry downloads, PyPI public download records, package metadata, and reverse-dependency data for tracked npm and PyPI packages.
About
AI-Buzz publishes AI industry research from dated public company data.
This page explains the public sources, source freshness, metric limits, research card review, and correction process behind that work.
AI-Buzz uses public sources when they can be mapped to a company and checked against source records. The source list changes only when the source can support dated company research.
npm registry downloads, PyPI public download records, package metadata, and reverse-dependency data for tracked npm and PyPI packages.
GitHub public repository metadata, stars, forks, commit activity, releases, and contributor activity for mapped repositories.
Hacker News Algolia results, including public discussion mentions and HN Who's Hiring comments when they can be mapped to a company.
Company websites, public source links, funding announcements, investor or news records, and manually checked company mappings.
Most package, code, discussion, and hiring metrics refresh on a daily cycle, with source lag when a provider has not settled recent days. Other datasets refresh weekly or monthly by source. Funding is published after verification. The as-of date is the date the displayed metric or cited source was observed for the research card; it is not the date a reader opened the page.
If a source is stale, incomplete, missing required metric evidence, or has a source link that cannot prove the metric, AI-Buzz should keep the older as-of date, show the limit, or avoid publishing the claim until the source can support it.
AI-Buzz combines public package and code metrics for consistent company context. Download metrics sum tracked npm and PyPI package streams attributed to a company. Company pages include metric dates and backing links when the source can be cited.
Fresh source data, clear company mappings, complete package or repository reads, anomaly checks, and manual review for high-impact claims matter most. Weak or stale data may appear as context, but not as the main claim in a research card.
Not every change becomes research. A company research card should make its metric, cited source, source link when available, and as-of date inspectable on the company page.
Public metrics do not prove revenue, market share, customer count, product quality, security posture, or private enterprise adoption. A metric is a dated public reading that needs source context.
npm and PyPI downloads can include CI, mirrors, testing, repeated installs, and short-lived automation. Package ownership and package-to-company mappings can also lag product changes.
Hacker News mentions reflect one technical community and query matching. Mention content can be positive, critical, repetitive, or unrelated to usage; small samples are noisy.
Hiring mentions, including HN Who's Hiring matches, do not represent the full labor market, actual headcount, or whether a role was filled.
GitHub stars, forks, commits, releases, and contributors cover public mapped repositories only. They miss private work, non-GitHub code, and production deployments.
Funding records are dated public events. They can lag announcements and do not prove revenue, current valuation, cash balance, or customer adoption.
Company mappings, funding records, cited sources, as-of dates, exact source links when available, and high-impact claims get manual review when automated checks are not enough. Missing evidence, stale sources, unsupported source links, or incomplete mappings can block publication. Missing an exact public source link by itself should not block a card when retained evidence, cited source, and as-of date are complete.
Factual errors, broken source links, missing cited sources, or disputed as-of dates can be reported from company pages or by emailing nick@ai-buzz.com. Include the affected company page, metric, source link if available, and the expected correction. Accepted corrections update the visible page data and the relevant as-of date.
The public changelog records data updates and source changes in the data changelog feed.
Research shows recent company research. Company pages show the metrics behind it.