Stitch vs Dex

Dex and Stitch both use AI to hire. Two things set them apart: where the candidates come from, and the intelligence that decides who is actually worth meeting. Dex is an AI talent agent focused on engineering: engineers sign up, talk to the agent about what they want, and get matched to companies. Stitch is not a marketplace, and it is not limited to engineering. It searches the live internet for the right person for any role you open, whether or not they have ever signed up to a platform, then scores everyone it finds with a model trained on your company.

Both differences compound. The pool decides who you can reach at all, and the strongest people are often the least likely to be on a platform. The scoring decides who, out of the hundreds of thousands Stitch considers, actually lands on your calendar.

The short version

• The pool. Dex matches you to engineers who signed up to its talent agent. Stitch reaches people anywhere on the internet, whether or not they have signed up to anything.
• Who is in it. An opt-in agent fills with engineers who are job searching. The strongest engineers are often heads-down and are not.
• Reach. Dex's pool is in the tens of thousands of engineers who have signed up, and far fewer are actively using it and keeping their profiles current. Stitch searches over a billion profiles live, and reads the real signals people leave in their work.
• Outcome. Dex introduces you to matched candidates. Stitch finds, scores, and books candidates from the whole market on your calendar.

At a glance

How each one actually works

Dex is a talent agent that engineers sign up to, then matches you to the ones in its pool. Stitch is not a marketplace. It trains a model of your company and keeps it at the center of the whole loop: learning who you are, searching the world live for the people who never signed up, and reaching out, with every result feeding back in.

The pool is everyone who signed up

Dex has focused on a single slice of the market, engineers and ML researchers, and built an AI talent agent that thousands of them have signed up to. For an engineer who wants to move, it is one way to be matched.

But the pool is still the engineers who signed up, and signing up is an act of looking. The engineers most teams are desperate for, the ones quietly doing the best work of their careers, are not browsing for their next role and did not join a talent agent. In engineering more than anywhere, the best signal is not a sign-up. It is the work itself: the repositories, the projects, the things they have actually built. A marketplace cannot see an engineer who is not in it.

Reach: a sign-up pool vs the whole market

Stitch does not rely on candidates coming to it. For each role, it deploys the Stitch Swarm: thousands of autonomous agents that gather candidate data live across the internet, wherever a person's footprint actually lives. For technical talent that often means source code repositories and project pages, exactly the places a heads-down engineer leaves a trail and exactly the places a sign-up pool cannot reach. Stitch searches across more than a billion profiles, none of whom had to join anything.

Scoring: a matching engine vs a model of your company

Dex matches within its pool. Stitch scores every profile it finds against your company specifically, with a custom model trained on your roles and your decisions, and surfaces only the top 0.1% for your role. The scoring engine was built by our founding team, ML engineers from Microsoft and Google with 17 AI patentsbetween us. It reads the real signals of quality in someone's actual work, whether that is what an engineer has built and shipped or a salesperson's quota attainment and President's Club finishes, rather than matching a shallow, self-reported profile. And because it considers hundreds of thousands of candidates per role, far more than any team could review by hand, the quality of that scoring decides who you ever get to see.

Outreach: introductions vs reaching the unreachable

Dex introduces you to candidates it has matched, which works when the engineer is in its pool. Stitch reaches the people who are not. It writes the outreach from a messaging template you approve, sends it from your team's own accounts, and books the interview on your calendar. Because it comes from a real, senior person at your company, the kind of note a founder or department head would send, rather than a platform, the people Stitch reaches reply. And because each candidate is scored against your bar, the interviews it books are high quality: customers accept around 90% of them.

It gets better with every interaction

A talent agent gets bigger as more engineers sign up. Your Stitch model gets sharper as you use it. Every scoring review, every declined meeting, every interview transcript, and every piece of post-interview feedback trains your custom model on what great looks like at your company. The model is sharper after your hundredth decision than it was after your first, and it is yours alone.

And it does not stop at the booking. Stitch helps your team prepare for each interview, supports you live as it happens, and helps with the decision afterward. Every interview signal then feeds back into your model, so the next role is sharper than the last.

So which should you choose?

If the engineer you want has signed up to Dex and is open to a move, Dex is a fast way to meet them.

If the engineers you want are heads-down and were never going to sign up to a talent agent, you have to go and find them where they actually are. That is what Stitch was built to do, for engineering and for every other role you define.

You can start a 14-day trial and see real candidates on your calendar before you decide. Most customers only pay on a successful hire.

FAQ

Dex is focused on engineers. Is Stitch as good for technical hiring? Yes, and that is exactly where it started. Stitch was built by engineers, our founding team holds 17 AI patents between us, and we built it for technical hiring first. It reads the signals engineers actually leave, including source code repositories and projects, and scores them with a custom model for your role. It worked so well that we now apply the same technology to any role you define, not just engineering.

Can Stitch reach engineers who never signed up to a platform? Yes. The Stitch Swarm finds candidates wherever their data lives, and outreach goes from your own accounts, so an engineer does not need to have joined Dex or heard of Stitch to land on your calendar.

Is a specialized talent agent higher quality than open search? A specialized pool sounds higher quality, but it is still only the people who signed up, and signing up is an act of looking. Narrowing to one niche does not change that: you get a tighter group of people who are on the market, when the people doing the strongest work are often not. Stitch searches the whole market and applies a custom model to surface the top 0.1% for your role, so you get reach and a high bar.