The whole job is one judgment call, made in seconds, on every inbound call: is this a real customer, or is this junk? Here's how Kathy makes that call, and why she gets it right far more often than a blocklist ever could.
Most spam tools work off a list of known-bad numbers. The problem is that spammers change numbers constantly and fake local ones to match your area code, so the list is always a step behind. Kathy doesn't rely on who's calling. She listens to what the caller actually says. A number she's never seen before means nothing to her. What the person on the line is trying to do means everything.
When a call comes in, Kathy picks up and finds out why the person is calling, the way a sharp receptionist would. Most callers reveal themselves in the first few seconds. A real customer gets to the point: they have a leak, they need a quote, they're calling about an appointment or a bill. A spammer is trying to sell you something, or it's a recording, or there's dead air.
Kathy passes a caller straight through when they sound like someone who needs your business. The signals are simple and human:
She stops the call when it carries the marks of a pitch or a bot:
This is the most important rule, and it's deliberate. Kathy is tuned to err on the side of passing calls through. If a caller is borderline, if she can't tell for certain, she puts them on to you rather than risk cutting off a paying customer.
Kathy isn't a black box. Every call she handles is logged in your dashboard with a recording and her decision. If she ever stops someone she shouldn't have, you'll see it, and you can add that number to your allow list in two seconds so it never happens again. Over time, that feedback makes her sharper for your specific business.
That's the difference between a filter and a decision. A filter blindly blocks. Kathy listens, judges, errs toward your customers, and shows her work.