2026-09-29
每日一文 · 长文精读
Presentation: From Consumers to Builders: Turning 200 of our Team into Agent Creators in 2 Weeks
从消费者到构建者：两周内将我们200名团队成员转变为智能体创造者
作者：Ben Maraney · InfoQ 原文
摘要：本文介绍了Forter公司如何在两周内让200名研发团队成员（包括非技术分析师）学会构建AI智能体。作者Ben Maraney分享了他们的方法：通过构建内部MCP服务器（Toolchain）集中管理工具，提供多个框架选择，并扫清组织障碍。他们让工具易于发现、理解和创建，从而快速扩展了工具库，并利用企业搜索工具Glean实现RAG功能。最终，团队在两周内从约20个工具扩展到近100个，证明了构建智能体可以很简单。
I was at QCon in London a few months ago, and in one of the breakout sessions I got talking to Krys Flores, who was one of the speakers.
几个月前，我在伦敦的QCon大会上，在一次分组讨论中与演讲者之一Krys Flores聊了起来。
I mentioned some work we've been doing around enabling people across R&D to build agents for ourself.
我提到了我们一直在做的一些工作，就是让整个研发部门的人都能为我们自己构建智能体。
She looked at me and she said, you know in my organization that seems impossibly far away.
她看着我说，你知道吗，在我的组织里，这似乎遥不可及。
What I realized was that mostly they were dealing with the hard parts about building agents.
我意识到，他们大多在应对构建智能体的困难部分。
I thought about what we'd done, and it hadn't felt that hard.
我想了想我们做过的事，感觉并没有那么难。
The truth is that what happened was we just sidestepped a lot of the hard parts about building agents.
事实是，我们只是绕过了构建智能体的许多困难部分。
What I want to do with this talk is to tell you that building agents does not need to be hard.
我这次演讲想告诉你们，构建智能体并不需要那么困难。
You can and you should make it easy.
你可以，也应该让它变得简单。
Early on in the company's development, we decided that every decision we made for merchants, no matter how high stakes, needed to be made in real time, which meant that we couldn't have humans involved doing manual review.
在公司发展早期，我们决定为商家做出的每一个决策，无论风险多高，都必须实时完成，这意味着我们不能让人工参与手动审核。
To do that well and precisely, we need a system which understands deeply both how fraudsters behave, but also both how normal customers, normal individuals behave.
为了做得又好又准，我们需要一个系统，既能深入理解欺诈者的行为，也能理解普通顾客、普通个人的行为。
The people who can do that research best are not the people in here, they're not engineers.
最能做好这种研究的人不是在座的各位，他们不是工程师。
They're analysts.
他们是分析师。
What they had in common when they joined Forter is that, largely speaking, they haven't really met code before.
他们加入Forter时的共同点是，大体上，他们之前从未真正接触过代码。
We needed to enable them to be able to do research and build logic that goes into our system.
我们需要让他们能够进行研究并构建进入我们系统的逻辑。
I was asked to be a founding member of an AI engineering team we were just setting up.
我被邀请成为我们正在组建的AI工程团队的创始成员。
What we got asked to do first was to run this two-week sprint hackathon thing in which people would get hands-on experience building their own agents so they could develop intuition about what AI does well and what it does less well, and to start them planning for this agentic future where also confidence is coming.
我们接到的第一个任务是举办一个为期两周的冲刺黑客马拉松，让参与者亲手构建自己的智能体，从而培养对AI擅长什么和不擅长什么的直觉，并开始为这个智能体未来做规划，同时建立信心。
We were told it needed to be for all of R&D, which included those analysts I mentioned before, which meant that it needed to be relatively easy to do even for folks who aren't that technical.
我们被告知这需要面向整个研发部门，包括我之前提到的那些分析师，这意味着即使对于不太懂技术的人来说，也必须相对容易上手。
We also got told that we had five weeks to do it in.
我们还被告知只有五周时间来完成。
This got us thinking very much about what is the absolute minimum we need to have a whole bunch of people building their own agents.
这让我们深入思考，让一大群人构建自己的智能体，我们需要的绝对最低限度是什么。
It took us to three things, tools, frameworks, and getting roadblocks out of people's way.
我们归结为三件事：工具、框架，以及扫清人们面前的障碍。
I'm going to break up the talk into those three things as well.
我也会把演讲分成这三个部分。
I'm going to try and convince you that with tools, building your own in-house MCP server is really high ROI and definitely worth doing.
我会试图说服你，在工具方面，构建自己的内部MCP服务器投资回报率很高，绝对值得一做。
I will try and convince you that you need to have more than one framework, and I'll explain why.
我会试图说服你需要拥有不止一个框架，并解释原因。
I'll give you some tips and tricks to get various roadblocks your organization might have out of your way.
我会给你一些技巧和窍门，来扫清你组织可能面临的各种障碍。
Before, just as the team was founded, my colleague Andy had put together this thing called Toolchain, which was an MCP server he was using to connect his Copilot at the time to various tools in his development workflow, Asana and some metric stuff and some Kubernetes stuff.
就在团队成立之初，我的同事Andy搞了一个叫Toolchain的东西，这是一个MCP服务器，他当时用它来连接他的Copilot与开发工作流中的各种工具，比如Asana、一些指标工具和一些Kubernetes工具。
What it did was it gave us this nice UI where people could go and see what tools were out there, what we had available, and then click into the tools and explore how they work and try them out.
它的作用是提供了一个漂亮的用户界面，人们可以进去查看有哪些工具、我们有什么可用的，然后点击工具探索它们的工作原理并试用。
Finding tools was suddenly easy because they were all in this one place.
寻找工具突然变得容易了，因为它们都集中在一个地方。
Once you found the tools you wanted, you would create a connection for your agent.
一旦找到你想要的工具，你就可以为你的智能体创建一个连接。
You just select the tools that it should have access to, because if you give an agent access to too many tools, it gets confused and it can do some bad things.
你只需选择它应该访问的工具，因为如果你给智能体太多工具的访问权限，它会感到困惑并可能做出一些不好的事情。
They would create the connection to the agents, choose the tools, do a little bit of configuration for them, and then it would spit out some connection details that you could then hand over to your agent so it would have access to the tools you wanted and only the tools you wanted.
他们会创建与智能体的连接，选择工具，进行一些配置，然后它会输出一些连接细节，你可以将这些细节交给你的智能体，这样它就能访问你想要的工具，而且只访问你想要的工具。
When the team was founded, we have about 20 or so tools in Toolchain.
团队成立时，Toolchain中大约有20个左右的工具。
Five weeks later when the sprint kicked off, we were close to 60.
五周后冲刺开始时，我们接近60个。
By the time the sprint was over, two weeks later, we had nearly 100.
两周后冲刺结束时，我们有了近100个工具。
Because it was all in this one repo, because there were all these examples, people found it super easy to add their own tools.
因为所有东西都在这个仓库里，因为有这么多示例，人们发现添加自己的工具非常容易。
That's not something you're going to do well or in a generic way in a two-week sprint.
这不是你在两周冲刺中能以通用方式做好的事情。
Yet you still need a way to retrieve context to your agents.
然而你仍然需要一种方式来为你的智能体检索上下文。
What I want to call out in particular is that we have this tool called Glean, which is this enterprise search tool that connects to a whole bunch of different data sources in our company.
我特别想提的是，我们有一个叫Glean的工具，这是一个企业搜索工具，连接了我们公司一大堆不同的数据源。
Could be things like Confluence or Asana or Jira or Slack or Salesforce, a bunch of useful data.
可以是Confluence、Asana、Jira、Slack、Salesforce之类的东西，一堆有用的数据。
They've been around for ages and they're really good at indexing that stuff, and they're really good at search on top of that.
它们已经存在很久了，非常擅长索引这些东西，而且在此基础上搜索也很出色。
阅读理解
1. What is the primary purpose of the Toolchain MCP server mentioned in the talk?
2. According to the talk, what is a common characteristic of the analysts at Forter?
3. Why did the team avoid building a generic RAG system during the two-week sprint?