Atlassian shouldn’t have become this interesting of a company.

It was started ten years ago, at a time when most of the major categories of enterprise software were saturated– even if they were saturated with lousy software everyone hated. After spending billions on stuff that barely worked, Corporations simply weren’t ready for a reboot. So the best option at building a business was filling in verticals with specialized offerings.

But good God, did Atlassian pick the right vertical.

It makes software for software developers that allows things like collaboration and bug tracking, and it’s announcing a new version of its core software today, JIRA 5.

Something like JIRA used to be confined to a few nerds on the periphery of corporate America. But today, not only is software a huge industry itself, it is eating the world. The line between what is a software compay and what isn’t is becoming hopelessly blurred, making Atlassian’s once geeky vertical core to nearly every single business trying to stay competitive. “There are more lines of code in a Ford car than in Twitter and Facebook combined,” says Atlassian President Jay Simons.

And the responsibility for making sure software works is intertwined throughout the corporate ladder, because software is so vital, whether it’s the core touchpoint with customers or the basic tools knowledge-worker employees need to function.

This hits home with me as a business owner. A few years ago, as an editor I would never have to think about anything but words and images. Today I have to know how to spot a rogue <div> tag, and have to track bug requests and fixes. Put another way, ten ago when Atlassian was founded someone like me would have had no need for JIRA. Today, I do.

The company has always been candid about just how well it’s doing: It has had a compound annual growth rate of 47% for the last five years, forty quarters of straight profitability, and did $103 million in annual revenue last year without a single salesperson. It has 20,000 active customers.

The top line numbers have been reported before, but the complete lack of a salesforce is almost unbelievable. Companies talk all the time about software that sells itself but most cloud and open source companies ultimately find they need sales people to get beyond $20 million a year or so in revenues. Even the most social, self-service and viral business software companies like Yammer are building sales forces.

Sound like a company that could go public? Exactly. We are hearing that Atlassian is already having those conversations with bankers. This will be yet another win for Accel, who gobbled up the company’s entire $60 million venture round back in 2010. It was a nosebleed valuation, and the largest enterprise deal Accel has ever done.

The new release of JIRA pulls more social features– like @mentions– into the software and has improved the API functionality. Atlassian is also announcing integration with thirty partners, including other hot enterprise up-and-comers like Box and Zendesk. Now, for instance, a customer’s complaint that relates to a bug fix might originate in Zendesk, and could be tracked and fixed in JIRA and then flow back into Zendesk when it’s resolved.

Strategically, it’s a smart direction for Atlassian. JIRA is already used by 70 of the Fortune 100, so the game is less about getting a foothold and more about expanding that foothold. Not only do the integration partners and added social features mean developers never really leave the program during their work days, but it further broadens the scope of who uses Atlassian’s software. More seats equal more money.

A lot of the young enterprise set is moving towards this play-nice API direction, rather than intentionally making their software harder to integrate ala the last generation of enterprise software kingpins.

But there may be some friction lurking. Right now there’s a feel-good API stew of these up-and-coming social enterprise players all wanting to support one another. That’s in keeping with the consumer Internet world, where people generally believe it’s not a zero-sum game and there is room for multiple players.

But in the enterprise world, where people pay for software, a land-war might develop between who wants to be the knowledge worker portal and who wants to be a mere API partner integrating into it.

[Accel is also an investor in PandoDaily.]