Loading...
Signal exists because the distance between what companies say about AI and what they actually do is measurable — and that measurement matters.
Every earnings call mentions AI. Every product launch is “AI-powered.” Every company is hiring a Chief AI Officer. But when you look at the actual engineering footprint, the hiring data, the product evidence — the picture is very different.
Signal was built to cut through the noise. We score companies on what the data actually shows — not what the press release says. The result is a single number and a verdict that tells you whether a company is genuinely building AI capabilities, or just managing perception.
We pull from seven public data sources — GitHub activity, job postings, SEC filings, funding rounds, press coverage, leadership profiles, and company websites. Each source is scored across six dimensions of AI maturity.
The flagship metric — the Narrative Gap — captures the delta between what a company says and what it ships. When that gap exceeds 25 points, it triggers a “Marketing-Led” verdict override. That's the theater detector.
The average Narrative Gap across all companies we've scanned is 31 points. Most companies talk about AI far more than they build it.
We score what's visible. If a company has incredible private AI labs but no public signal, they'll score low — and that tells you something too.
Every company claims to be "AI-first" now. We measure the gap between narrative and substance — the Narrative Gap is the feature, not the bug.
Every score includes confidence levels and source citations. We show our work so you can disagree with the data, not the methodology.
Self-assessed AI maturity is marketing. We pull from GitHub, SEC filings (public cos), Crunchbase, job postings, company websites, and funding data — sources that carry consequences for inaccuracy.
Signal is not affiliated with any company we score. No advisory relationships, no consulting deals, no conflicts of interest.
Six dimensions, seven data sources, weighted scoring with confidence intervals. We built Signal because the existing landscape of AI assessments was built on vibes.
Free scans for everyone. The overall score and verdict are always free. We charge for the deep analysis — not the headline.
Signal is built on Next.js and Supabase. Every report starts with parallel data collection from seven public sources — including direct website scraping via Firecrawl — followed by structured AI evaluation against scoring rubrics for each dimension.
Reports are generated asynchronously via Inngest and cached for 24 hours. The scoring model is chosen for its ability to evaluate evidence against nuanced rubrics with confidence levels.
We're continuously improving the scoring model and expanding data source coverage. If you have feedback on a specific report, we want to hear it.
Analyze any company and see the Narrative Gap for yourself. Free, instant, no signup required.
Analyze a Company — Free →