How we use AI

The short version

We measure AI answer engines for a living, so yes — we use AI heavily, and it would be strange if we didn't. Software and AI agents run our measurement pipeline and help draft materials. A human analyst reviews every client-facing result, writes the final report, and takes every call. We are responsible for everything we publish and deliver.

The measurement itself

Your Snapshot works by asking real AI engines real questions. That means your business name, website, and market context are sent as queries to AI services (such as Anthropic's and Perplexity's APIs) during measurement — the same way any buyer's question about your market would reach them. We send public information only.

Where AI helps behind the scenes

• Running the checks: scripted, deterministic tools do the fetching, classifying, and scoring; AI models help generate candidate buyer questions and classify answers. Every number in a report traces to logged evidence.
• Drafting: AI tools help draft report prose and site copy. Nothing ships without human review and editing.

What we do not use AI for

• Judgment calls: which questions represent your real buyers, what the results mean, and what to fix first — a human analyst decides, every time.
• Case studies and claims: the results we publish are real and traceable. We never fabricate a statistic, a testimonial, or a result — one fake number would end this business, and we know it.
• Support: mail to hello@buyermining.com is read and answered by a person.

Why we publish this

Our whole pitch is that AI answers now shape who gets hired — so you deserve to know exactly where AI sits in our own work. The full measurement methodology is on the Method page.