Building AI Agents for Everyone

Building AI Agents for Everyone

Gumloop 创始人揭秘:如何让全员构建可控 AI 智能体

YC Root Access3 位订阅者

70 分钟阅读3 次阅读AI

【AI 前沿】从工作流到智能体:Gumloop 如何赢得 Shopify 与 Instacart 信任 Gumloop 联合创始人 Max 和 Rahul 在 YC 访谈中分享产品演进历程。他们讲述如何从早期工作流自动化转型为企业级 AI 智能体平台,通过 1100 次客户电话迭代出满足 IT 安全合规需求的解决方案。视频深入解析了透明定价策略、多云推理架构如何赋予企业“智能体主权”,以及为何“理解任务的人”应亲自自动化工作。这对希望在企业内部落地 AI 的团队极具参考价值。 🎙️ 主讲人介绍 Max Brodeur-Urbas:Gumloop 联合创始人,负责产品战略与企业客户拓展,曾四次申请 YC 并最终入选。 Rahul Behal:Gumloop 联合创始人,主导技术架构与 SDK 开发,关注智能体在企业工具中的集成与控制。 Gustaf Alströmer:Y Combinator 普通合伙人,主持本期 Fireside 访谈,引导嘉宾分享创业心得。 核心议题:AI 智能体平台演进、企业级安全与合规、透明定价策略、员工自主自动化、多云推理架构。 关键词:Gumloop、AI Agents、Enterprise AI、Workflow Automation、YC W24

章节 1:Gumloop:面向所有团队的 AI 智能体构建平台

📝 本节摘要:

本节介绍了 Gumloop 作为 AI 智能体构建平台的定位与功能。联合创始人 Max 和 Rahul 解释了产品如何帮助 Shopify 等大型企业让全员构建并共享智能体。他们回顾了从工作流自动化到智能体平台的演变,演示了通过 UI 创建智能体并将其集成到 Slack 等工具的流程,并强调了其作为连接层而非替代层,以及通过 SDK/API 支持企业内部工具开发的能力。

[原文] [Gustaf Alströmer]: Today I'm here with Max and Rahul from Gum Loop. thanks for coming back to IC.

[译文] [Gustaf Alströmer]: 今天我和来自 Gumloop 的 Max 和 Rahul 在一起。谢谢你们再次来到 YC。

[原文] [Max Brodeur-Urbas]: Thank you for having us.

[译文] [Max Brodeur-Urbas]: 感谢邀请我们。

[原文] [Gustaf Alströmer]: I'm really excited for this conversation. You guys are one of the most successful yet kind of unknown companies from W24. Tell us what Gumloop is.

[译文] [Gustaf Alströmer]: 我对这次对话感到非常兴奋。你们是 W24 批次中最成功但相对低调的公司之一。请告诉我们 Gumloop 是什么。

[原文] [Max Brodeur-Urbas]: Gumloop is the agent builder that a lot of companies like Shopify, Gusto, and Instacart use so that everyone can build agents and share them across the business while it can manage and control that data. So kind of like the holy grail of agents. This is what everyone is talking about, what they want to actually build. This is probably one of the more common applications we get from YC today, but you actually built it and you're deployed in many of the biggest companies.

[译文] [Max Brodeur-Urbas]: Gumloop 是一个智能体构建器,许多像 Shopify、Gusto 和 Instacart 这样的公司都在使用它,以便每个人都能构建智能体并在整个业务中共享,同时平台可以管理和控制相关数据。这可以说是智能体领域的“圣杯”。这是大家一直在谈论、想要真正构建的东西。这可能是我们在 YC 今天遇到的更常见的应用之一,但你们真的把它做出来了,并且部署在许多最大的公司中。

[原文] [Rahul Behal]: Yeah, it's definitely a crowded space because it feels like obvious that someone has to solve this problem. There's definitely a bunch of players that are doing a good job, but we're just trying to

[译文] [Rahul Behal]: 是的,这确实是一个拥挤的领域,因为显然必须有人来解决这个问题。肯定有很多玩家做得不错,但我们只是试图

[原文] [Gustaf Alströmer]: be one of the leaders. Yeah. And you applied to YC for Winter 24. You called it Agent Hub at the time.

[译文] [Gustaf Alströmer]: 成为领导者之一。是的。你们申请了 YC 2024 冬季批次,当时你们叫它 Agent Hub。

[原文] [Rahul Behal]: and then I switched to gum loop.

[译文] [Rahul Behal]: 然后我改名为 Gumloop。

[原文] [Gustaf Alströmer]: Yeah.

[译文] [Gustaf Alströmer]: 是的。

[原文] [Gustaf Alströmer]: What was this product when you applied?

[译文] [Gustaf Alströmer]: 你们申请时这个产品是什么样的?

[原文] [Rahul Behal]: Yeah, it is ironic that we called ourselves Agent Hub back then because now we are an agent hub, but there was a bunch of... everyone thought we were a real estate agent company because agents weren't common back then.

[译文] [Rahul Behal]: 是的,有点讽刺的是,我们当时叫自己 Agent Hub,因为现在我们确实是一个智能体枢纽,但当时很多人以为我们是一家房地产中介公司,因为那时候“智能体”这个词还不常见。

[原文] [Rahul Behal]: back in the like back then

[译文] [Rahul Behal]: 就是那时候。

[原文] [Rahul Behal]: But we were a workflow automation platform. So same mission of automating tasks, but the models weren't good enough back then. So we just made a visual builder and people could define their workflow step by step. Then we added agents, and that was kind of like the hockey stick moment where anyone could build it on the platform and usage skyrocketed. But same mission since the beginning.

[译文] [Rahul Behal]: 但我们当时是一个工作流自动化平台。使命同样是自动化任务,但那时的模型还不够好。所以我们做了一个可视化构建器,人们可以一步步定义工作流。后来我们加入了智能体,那有点像“曲棍球杆时刻”,任何人都可以在平台上构建,使用量随之飙升。但使命从一开始就没变。

[原文] [Gustaf Alströmer]: How does Gumloop work? How would you actually use it?

[译文] [Gustaf Alströmer]: Gumloop 是如何工作的?你实际上会怎么使用它?

[原文] [Rahul Behal]: Yeah. So basically you can go onto the platform and then create an agent and add...

[译文] [Rahul Behal]: 是的。基本上你可以进入平台,然后创建一个智能体并添加...

[原文] [Gustaf Alströmer]: How do you create it?

[译文] [Gustaf Alströmer]: 你是怎么创建它的?

[原文] [Rahul Behal]: You create an agent in the onboarding, but you can also just click the 'Create Agent' button and then you're presented with the standard chat interface. Then you have a set of connectors that you can pick and choose from all the apps that you work with. You can start talking to the agent right away to get work done, or you can add the agent to your Slack workspace, Microsoft Teams, wherever you work. We kind of bring the agents to you.

[译文] [Rahul Behal]: 你可以在入门引导中创建智能体,也可以直接点击“创建智能体”按钮,然后你会看到一个标准的聊天界面。接着你有一组连接器,可以从你使用的所有应用中进行选择。你可以立即开始与智能体对话以完成工作,或者将智能体添加到你的 Slack 工作区、Microsoft Teams 或你工作的任何地方。我们基本上是把智能体带到你身边。

[原文] [Gustaf Alströmer]: What's the most like what's some good examples of common agents that people make once they have the your UI?

[译文] [Gustaf Alströmer]: 一旦有了你们的 UI,人们制作的一些常见智能体有哪些好的例子?

[原文] [Rahul Behal]: They tend to be team-specific. So the sales team at pretty much all of our customers, for example, will have a sales agent that's connected to their call recording software, their CRM, a bunch of enrichment tools, web search tools, and it has a bunch of skills for how they like to do business and how they want to manage opportunities and analyze calls. It becomes like a co-worker for the sales team that the sales team uses and relies on. So they're pinging it in Slack every day. They're CCing it on email threads. They're having it set up next steps, clean and deduplicate accounts, like all of the tedious work that a salesperson would be doing. They built an agent to do that for them, and it's like their little co-pilot.

[译文] [Rahul Behal]: 它们往往是针对特定团队的。例如,我们几乎所有客户的销售团队都会有一个销售智能体,它连接到他们的通话录音软件、CRM、一堆数据丰富工具、网络搜索工具,并且具备一系列技能,用于他们希望开展业务的方式、管理商机和分析通话。它变成了销售团队使用并依赖的“同事”。所以他们在 Slack 里每天都会@它,在邮件线程中抄送它,让它设置后续步骤、清理和去重账户,就像销售人员会做的所有繁琐工作一样。他们构建了一个智能体来替他们做这些,它就像他们的小副驾。

[原文] [Gustaf Alströmer]: Does it replace the software tools that they are currently using, or is it kind of interacting with them? Like how does it work with, say, Salesforce or something like that?

[译文] [Gustaf Alströmer]: 它是取代了他们当前使用的软件工具,还是与它们进行交互?比如它是如何与 Salesforce 之类的东西一起工作的?

[原文] [Rahul Behal]: Yeah, we like to be the connector between all of your tools. So sometimes it actually does replace vertical SaaS you might have procured. For a lot of our customers, they're canceling contracts to just use an agent instead. But often times we're like the gum between your different data connectors. So you're using all the tools that you love, but you're bringing them into one place and giving it to this little co-pilot that can just do stuff for you in the background.

[译文] [Rahul Behal]: 是的,我们喜欢做所有工具之间的连接器。所以有时候它确实会取代你可能采购的垂直 SaaS。对于我们的许多客户来说,他们正在取消合同,转而只使用智能体。但很多时候,我们就像不同数据连接器之间的“胶水”。所以你在使用所有你喜欢的工具,但你将它们汇聚到一个地方,交给这个小副驾,它可以在后台为你做事。

[原文] [Rahul Behal]: A lot of companies are developing internal tooling. Their own engineering teams are building agents, and we kind of fit right into that ecosystem as well. Like we provide a pretty robust SDK and API. So pretty much anything you can do through the UI in Gumloop, you can do via API or SDK, and we have hundreds of integrations exposed via MCP servers. And so a lot of our customers are building internal tooling and infrastructure using Gumloop as a base layer for their AI.

[译文] [Rahul Behal]: 许多公司正在开发内部工具。他们自己的工程团队正在构建智能体,我们也正好融入那个生态系统。我们提供了相当健壮的 SDK 和 API。所以在 Gumloop 中通过 UI 能做的几乎任何事,你都可以通过 API 或 SDK 完成,并且我们通过 MCP 服务器暴露了数百个集成。因此,我们的许多客户正在使用 Gumloop 作为其 AI 的基础层来构建内部工具和基础设施。

  1. 章节 2:从工作流自动化到企业级 AI
  2. 章节 3:赢得企业客户的关键
  3. 章节 4:创始人相识与加入YC
  4. 章节 5:Gumloop 内部智能体应用与产品市场契合
  5. 章节 6:拿下 Shopify 与重新思考定价
  6. 章节 7:融资策略与保持精干的局限
  7. 章节 8:让员工自动化自己的工作
  8. 章节 9:构建企业可完全掌控的 AI 智能体
  9. 章节 10:IT 决策者的角色与旧金山的转折

正在加载后续章节…