Choosing an accelerator or incubator for your new startup is a lot like choosing where to go to college. The right entrepreneurship support organization (ESO) will introduce you to the people, investments, and ideas that will launch your company. The wrong ESO will waste your time and money.

Unfortunately, out of the roughly 2,500 ESOs in North America, most founders only know a handful. 

As a result, founders are likely to apply to YC (10,000 quarterly applications, 1% admittance rate), a local TechStars program, and whichever program was recommended by a friend. This is like only applying to MIT, Harvard, and the community college your cousin went to. Meanwhile, there might be dozens of lesser-known ESOs looking to support companies in your specific region or sub-industry.

That’s why I’m excited to present Atlas, a new feature on Passport that’ll find the best programs for your specific startup. Enter your startup’s website, tell us what you’re looking for, and Atlas returns a short list of accelerators, incubators, and pitch contests that are a real fit, including ones you’d never have found on your own.

What Opportunity Atlas Does

Atlas lives on the Passport homepage and is free for anyone – Passport account or not – to use.

Here’s what it asks for:

• Your company’s URL, which Atlas uses to pull context on your business
• Whether you’re looking for accelerators, incubators, pitch contests, or all three
• What funding stage you’re at

From there, Atlas gets to work. It compares your startup against a database of more than 1,000 accelerators, incubators, and pitch contests that are actively seeking companies right now, and ranks the best matches for you, with a deliberate bias toward smaller, more specific programs that founders often don’t know exist. You get up to six recommendations, an explanation of why each one made the list, and a direct link to apply.

That bias toward the smaller, less obvious programs is deliberate. The big-name accelerators already get thousands of applications, so a match there is a long shot no matter how good your startup is. A regional fintech accelerator with 40 applicants and a mandate to fund exactly your kind of company is a much better use of your time.

Why We Built This

I’ve spent the past five years working alongside more than two dozen ESOs, and tens of thousands of startups through OneValley’s Passport platform. Across all of that, one pattern kept showing up: startups and the programs built for them are having real trouble finding each other.

On the founder side, most people know the big names: YC, Techstars, maybe whatever program their university runs. Past that, the options are a Google search and a lot of guessing. Meanwhile, plenty of accelerators and incubators, some very well funded, are actively looking for companies exactly like theirs and going largely unseen.

On the program side, it isn’t much better. ESOs are wading through inconsistent, unverified information trying to figure out which startups are legitimate, which are at the right stage, and which are worth their time. It’s a little like a bad online dating experience, where half the profiles don’t match reality.

How We Built It

We didn’t start from scratch. In 2024, we worked with Seekr to launch Haystack, an AI shopping assistant for enterprise software, and came away impressed by how they approached explainability and auditability: making it possible to actually trace and fix why an AI system produced a given output, rather than treating it as a black box. When Seekr released SeekrFlow, their platform for deploying and maintaining AI products, it was a natural next step to build Atlas on top of it.

The hardest product question wasn’t the matching itself. It was figuring out how to learn enough about a startup to make a useful recommendation without putting founders through a long, painful intake form.

We tested three approaches. Extracting information from pitch decks sounded promising, but founders were understandably reluctant to share them, and the quality of what we pulled out was inconsistent. Connecting to sources like Crunchbase and LinkedIn was expensive and, just as often, out of date or incomplete. What actually worked, better than either of the more complex options, was well calibrated crawling of a startup’s own website. 

From there, the work became a matter of continually improving the accuracy and speed of the underlying matches, which is still ongoing.

Under the hood, Atlas runs on three coordinated AI agents. One continuously searches for and classifies new opportunities. A second researches your startup from your website to determine what you do and what you’d likely be eligible for. A third combines the two to generate your recommendations.

Who Atlas Is Built For

Atlas is built first for founders: specifically, pre-seed, seed, and Series A companies looking for accelerators, incubators, or pitch competitions. It’s especially useful if you’re outside Silicon Valley, where the well known programs are harder to access and the regional or industry specific ones are much harder to find on your own.

It’s also useful on the other side of the table. ESOs running accelerators, incubators, or competitions get their programs surfaced to relevant, matched startups automatically, rather than relying entirely on founders finding them through word of mouth or a lucky search. If that’s you, there’s a separate path to get your program included in Atlas’s database, which we’ll cover in a follow up post.

Try It

Using Atlas is simple: share your startup’s URL, tell us your funding stage, and let it do the work.

Right now, Atlas works the same whether or not you’re signed into Passport. That’s the next thing we’re changing. Our near term roadmap connects Atlas to the rest of Passport, so signed in users can generate and save as many matches as they want, get notified when a particularly strong match shows up, and see their recommendations improve the more they use the platform.

We built Atlas because we think founders deserve better than a Google search and a list of the same five names everyone already knows. If you’re building something outside the spotlight, there’s very likely a program out there looking for exactly you. Go find out which one.