Pillar · the spear
AI-Readiness Detection
AI-readiness is an infrastructure question, not a marketing one. Bassethound fingerprints the AI stack a company runs and returns a verdict you can act on: shipping, experimenting, or none.
The unclaimed axis. Everyone else reads the homepage for AI mentions. Bassethound reads the stack a company actually runs.
What AI-Readiness Means
What AI-readiness is, how to measure it, what an AI-readiness score captures, and which signals actually show a company is serious about AI.
Explore cluster →Detecting the AI Stack
Fingerprinting model providers, SDKs, orchestration frameworks, vector stores, and LLM observability from the outside.
Explore cluster →llms.txt and MCP Endpoints
What llms.txt is, standard versus technical llms.txt, detecting MCP endpoints and .well-known AI files, and why these machine-facing signals matter.
Explore cluster →Shipping vs Experimenting vs None
How to tell whether a company runs AI in production, only dabbles, or ignores it, and how the verdict is scored.
Explore cluster →AI-Readiness as an ICP Signal
Turning AI-readiness into a buying signal: finding companies building with AI, using readiness as an ICP filter, and qualifying prospects by their AI stack.
Explore cluster →Perspectives & resources
Frequently asked questions
What is AI-readiness detection?
It is identifying the AI infrastructure a company runs from the outside (model providers, MCP endpoints, vector stores, orchestration frameworks, and LLM observability) and scoring how far into production that AI is.
How is Bassethound's AI-readiness different from the shallow version?
Most tools read the homepage for AI mentions and check for an llms.txt. Bassethound fingerprints the actual backend AI stack and returns a shipping, experimenting, or none verdict, delivered over MCP.
Why does AI-readiness matter for sales?
A company shipping AI in production is a qualified buyer for AI infrastructure, tooling, and services. The AI-readiness verdict is an ICP filter you can act on.
Sniff a domain.
Run sniff_domain on any site and read its five-layer dossier in one call.