Matt Lawler
① 他说过的话
8 条
如果你想要在规模化下交付更好的客户体验，你就不能成为良好客户体验的瓶颈。
指向原始笔记的链接
If you want to deliver a better customer experience at scale, you can’t be the bottleneck to having a good customer experience.
—— Matt Lawler · [04:07]
于是我们「造」了另一个团队成员而不是去雇一个，我们给他起名叫 Joey。
指向原始笔记的链接
So we built another member of our team rather than hiring one, and we named him Joey.
—— Matt Lawler · [05:18]
实际上有几次，我们实时监控一场对话，发现了一个 bug，部署了修复，然后在会话的剩余时间里它就被验证有效了，而那个客户甚至不需要知道我们在幕后发布了那个修复。
指向原始笔记的链接
We’ve actually had times where we’ve actually monitored a live conversation, caught a bug, deployed a fix, and then it’s been proved for the rest of the session, and that customer doesn’t even have to know that we shipped that fix behind the scenes.
—— Matt Lawler · [07:42]
所以，在部署这个构建的第一周内，我们就把端到端解决率从 10% 提升到了 80%，而且用的是相当朴素的实现。
指向原始笔记的链接
So we went from 10% to 80% end-to-end resolution rate in just the first week of deploying this build with a pretty naive implementation.
—— Matt Lawler · [08:09]
而且我们总共只花了大约每月 700 美元的 token 和基础设施成本就做到了这一切。
指向原始笔记的链接
And we did that all for around $700 a month in both token and infrastructure costs.
—— Matt Lawler · [08:17]
理解你的客户在构建什么，最好的方式就是你自己真正去构建他们的同样产品。
指向原始笔记的链接
The best way to understand what your customers are building is to actually build their same product yourself.
—— Matt Lawler · [10:11]
我记得那大概是一个 30000 行的 ClaudeMD，里面全是各种护栏，以及关于他应该如何与客户打交道的建议。
指向原始笔记的链接
I think it’s like a 30,000 line ClaudeMD of all these guardrails and advice on how he should be operating with customers.
—— Matt Lawler · [11:37]
所以这里面有一种元层面的东西：当你在和 Joey 聊如何构建语音智能体 API 时，你已经在使用它了，你已经亲眼看到你的产品体验会是什么样子。
指向原始笔记的链接
So there’s kind of a meta layer there where if you’re talking to Joey about how to build a voice agent API, you’re already using it, and you’re already seeing exactly what your product experience could look like.
—— Matt Lawler · [14:12]
② 出现在这些集
1 集
- 《一个AI客服解决了80%的工单，成本每月700美元》 — 作为主持
③ 他谈到的
点进去有真内容 —— 本页主要出口
Assembly AI · Joey · 前向部署工程师 · 语音智能体 · Claude Agent SDK · RAG · ClaudeMD · Railway · 护栏 · 内部试用