TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Show HN: Product recommendations powered by human experts and ML, fast and easy

3 点作者 goldenjm将近 2 年前
Hey HN!<p>Joe here, I’m a second time founder based in Seattle. I wanted to share a new site I’ve been building called PerfectRec. We use a combination of human expertise and ML to generate personalized electronics recommendations based on user answers to straightforward questions.<p>Google is getting worse for product discovery. Amazon is full of fake reviews. And many of the review sites I used to trust seem to be going downhill.<p>As the go to “gadget guy” in my family, I wanted to build a site I could point people to that would give them high quality product recommendations without requiring expert knowledge or knowing how to navigate Reddit. Instead of it taking hours to research a complex purchase, PerfectRec takes just minutes.<p>How it works: We find the most passionate product experts on the internet and have them work with our ML team to design, label and train a recommendation model.<p>Right now, we cover phones, TVs, laptops and earbuds in the US market, though we are planning to expand to cover more products.<p>I’d love to hear what you think, since the site is meant for highly tech savvy people too. Would you recommend it to a friend or family member? What would you like to see us add or improve?<p>Thanks!<p>Joe

暂无评论

暂无评论