HackerNews中文版

简而言之：我们正在开发 Webhound（https://webhound.ai），这是一个能够从网络构建自定义数据集的 AI 代理。我们正在寻找测试用户，在测试阶段可以免费无限制使用该产品。 ---------- 大家好！在过去两年里（自从从 YC 毕业以来），我们一直在搜索领域开发产品，最近几个月我们致力于构建 Webhound，这是一个专门用于数据收集和聚合的深度搜索代理。它可以在网上搜索、访问网站和观看 YouTube 视频。它拥有工作记忆和内部记事本，用于制定和执行计划，并且可以并行部署一组深度搜索工作代理，以回答问题、提取信息和分析搜索结果。我们正在寻找测试用户来帮助我们解决产品中的任何问题。如果您对此感兴趣，请在下面留言或填写这个谷歌表单，我们将向您发送邀请代码以供试用：https://forms.gle/7Wip9toggKDRX2zAA 另外，这里有我的联合创始人 Theo 的一段视频，提供了关于如何使用 Webhound 及其工作原理的更深入指南：https://www.youtube.com/watch?v=JGtQlgULYFw

查看原文

TL;DR: We’re working on Webhound (https://webhound.ai), an AI agent that builds custom datasets from the web. We’re looking for beta users who will receive free unlimited access to the product while it is in beta.----------Hey everyone! After working on products in the search space for the past two years (since graduating from YC), we’ve spent the past few months building Webhound, a deep search agent that is purpose-built for data collection and aggregation.It can search the web, visit websites, and watch YouTube videos. It has a working memory, an internal notepad which it uses to make and execute on plans, and a fleet of deep search worker agents which it can deploy in parallel to answer questions, extract information, and analyze search results.We’re looking for beta users to help us iron out any kinks in the product. If any of this sounds interesting to you, please either drop a comment below or fill out this google form and we’ll send you an invite code to try it out: https://forms.gle/7Wip9toggKDRX2zAAAlso, here’s a video from my co-founder Theo giving a more in-depth guide to how to use Webhound and how it workshttps://www.youtube.com/watch?v=JGtQlgULYFw

寻找Webhound的测试用户，这是一款能够从网络上构建数据集的AI代理。