The fair questions.

Cold lists are random strangers. Buddy only shows you people who have already raised their hand publicly — they made a post, complained about a competitor, or hired for a role your agency could fulfill. Every lead is something a human wrote this week, with real buying intent.

Heavy. You set the brief — business, ICP, tone of voice, channels, daily volume. You see every send in the live activity feed with full message text. You can pause or kill the agent in one click from the sidebar. What you don't do is click-to-approve each individual message — the whole point of an agent is that you don't have to.

Every post goes through three filters — keyword + spam check, relevance read, and a final scoring pass. Each lead ends up with a 0–100 intent score and a type tag (Direct Request, Problem Match, or Competitor). You see the score front and center.

Two places. It watches public posts on Reddit and Facebook for people signalling they're ready to buy, and it scans public job boards for companies hiring for a role your business could fulfill. Then it reaches out where they'll actually see it — a Reddit or Facebook DM, a WhatsApp, or an email from your own inbox.

No — it's built brake-first. Buddy sends from your own accounts at a careful, human pace, well within what the platforms expect. It only ever reaches out to leads you've saved, never random strangers, never deletes anything, and you can pause it in one click. We tune the safety limits conservatively and keep them that way.

No. Buddy only uses public APIs and public posts. We don't scrape emails, touch private groups, or violate platform terms. Every lead links to the original public post so you can reply where the person already is.

No. The demo walks you through it end to end. After that, paste your website URL or describe your business in one sentence — Buddy figures out your ideal customer, the right keywords, and where to listen. Setup takes under 60 seconds.