Job Search with Metaphor

Finding a job is way harder than it should be. Tools like LinkedIn, Handshake, or Google are supposed to solve this problem, but they’re filled with too many noisy results to actually be useful.

Here's how I used AI to find hundreds of hidden job listings in less than 5 minutes.

Metaphor is a new search engine — built from scratch using recent advances in AI — that is ideal for finding lists of high-quality results.

In 5 minutes with Metaphor, I created this list of 1000+ tech job listings that matched exactly what I was looking for.

At a high level, Metaphor is a search engine that understands your query. So, when I searched “ML internships for new grads in San Francisco” here’s what I got:

And, by filtering for only things that were posted recently, I made sure that the positions were new and not-filled.

But, there’s actually an even better ✨magical✨ way to take advantage of Metaphor. I can just paste a job posting I like and get similar ones.

For example, let's say I found a great-looking OpenAI Internship and now I want to find hundreds more just like it. If I paste the link to the internship I already found, Metaphor will return a list of similar and relevant positions.

Pretty cool right.

More than just jobs

Job search is really just one use case of Metaphor. Metaphor is a search engine built using novel representation learning techniques — ie it’s a novel way to search and research that’s very different from old-school search engines like Google.

For example, Metaphor excels at finding similar things.

• shopping: if you want a similar (but cheaper) shirt, paste a link to your shirt and it’ll give you hundreds like it.
• research: paste a link to a research paper to find hundreds of other relevant papers
• startups: if you’re building a startup, find your competitors by searching a link to your startup

As a bonus, here's a list of 1000+ tech internship listings I found.

hope you enjoyed this read,

Sarah Chieng

PS: we're also hiring! Check out roles with Metaphor here.