LinkedIn's first investment is data-infrastructure startup Confluent

Business-oriented social networking company LinkedIn has entered the venture funding world, backing a brand new big data startup called Confluent that has spun out of LinkedIn.

See, the technology Confluent is focused on — the Apache Kafka open-source project for immediately pushing data into all the places where it needs to be — came out of LinkedIn in response to the company’s engineering needs.

The people behind Confluent are some of the key people behind Apache Kafka, and LinkedIn loves Apache Kafka and the people behind it. So it makes sense for LinkedIn to invest in Confluent’s new $6.9 million round as Confluent goes out to help companies make the best of Apache Kafka as a contemporary enterprise-messaging tool for the big data world.

Here’s how the deal went down:

A couple of months ago, Jay Kreps, the co-creator of Apache Kafka, told senior vice president of engineering and operations Kevin Scott, that he’d chosen to start-up Confluent with colleagues Jun Rao and Neha Narkhede.

“[Scott] was actually pretty excited,” Kreps told VentureBeat in an interview today. “He said, ‘Hey, this is awesome, this is really what we like to see, you know, happen. … I’d like to find a way for us to invest, and so let me talk it over with our CEO, Jeff [Weiner].’ He did that, and the next day, he said, ‘Yeah, we’d be interested. Let’s figure out what makes sense.'”

Kreps and the Confluent team got help assembling a funding pitch. Then it went around telling its story to investors across Silicon Valley, he explained.

The result: Benchmark Capital is leading the round in the month-old startup. Data Collective is also participating, along with LinkedIn.

But don’t get the wrong idea. LinkedIn isn’t about to start a venture arm.

“We are not disclosing the specific amount LinkedIn’s investment, are not setting up a venture arm, and have no additional plans to make investments in other startups,” a LinkedIn spokeswoman wrote in an email to VentureBeat.

It’s just that for LinkedIn, the one and only Apache Kafka startup — for now — is fairly important for LinkedIn. And vice-versa.

“We want to continue working with them really closely,” Kreps said. “There’s a bunch of really talented infrastructure engineers there. We also roll out a lot of changes in new software very early on, often before it’s released in open-source. … That kind of real-world production setting is really essential.”

In that sense, Confluent’s relationship with LinkedIn is sort of like Hadoop distribution provider Hortonworks’ relationship with Yahoo, where Hadoop was originally created. Mountain View, Calif.-based Confluent has an origin story to tell and a great proving ground for its tools right from the beginning.

At the same time, Confluent benefits from the open-source community around Apache Kafka beyond LinkedIn, which includes thousands of companies, like Netflix, Pinterest, Uber, and Verizon, Kreps wrote in a blog post today. That community could provide a talent pool and a place to start looking for customers.

And people do want support for Apache Kafka. Kreps is certain of that much.

“Through open source, we were interacting with more companies who were using Kafka for these bigger and bigger use cases,” he said. The companies wanted to somehow get support from LinkedIn engineers. Thus the light-bulb moment about a year ago, when Kreps thought it might not be such a bad idea to start a company.

“We had a lot of really good phone calls really looking for commercial support,” Kreps said. “It wasn’t a huge leap to kind of see that.”