Loading page...Загружаем страницу...正在加载页面...
AtlasRepo

I built an after-hours front desk that runs on the shop's own laptop, in the owner's voice, with no cloud AI

说明与用途

为小型服务公司创建本地语音接待,处理下班后的电话,判断紧急程度并生成摘要卡片给店主查看。

问题

Small trade shops cannot staff a phone after hours and existing solutions are expensive or impersonal.

结果

The system achieved 95% accuracy on urgency judgment, 100% on emergency flagging, 98.75% on field extraction, and 90% on complete scenario accuracy across 20 scripted calls.

工作流

  • 1. Call comes in
  • 2. Desk answers and speaks in the shop owner’s voice
  • 3. Desk asks for problem, name, phone, address, requested time
  • 4. Desk judges urgency against shop’s written rules
  • 5. Desk records summary card at hang‑up
  • 6. Card appears in owner’s morning review list

证据

The desk was not tuned against these scenarios first. They were run once each, as written.

来源

Urgency judged right 19 of 20 (95%)

来源

The desk answers out loud, asks what the problem is, takes a name, a callback number, an address, and a requested time

来源

  • 记录日期: 2026-10-14T03:41:09.656Z