by Alex Konrad
A recommended film or former colleague can be accepted or ignored in just heartbeats. But the data required to get them in front of a user’s eyeballs is so vast, it’s hard to fathom. Linkedin, for example, is powering 300 billion events in which data is used each day, or about 1,000 for every American alive. At any given time, Uber tracks thousands of rides and drivers to match them up and then route the best journey at once; Cisco scans thousands of traffic intersections with many more thousands of cars.
So all those companies use a new approach to how they handle that data called a stream. Continuously flowing and never-ending, the stream passes data by all the outlets that might need it without them having to wade in and find an individual piece. That saves valuable time (you want that movie title near instantaneously after logging in, made as accurate as possible from all your past behavior) while allowing the same data to be used in multiple places in rapid succession, as opposed to waiting for it to be return like a checked out library book.
That stream is the creation of a team of former LinkedIn employees who built an open-source technology called #apache APA -4.02% Kafka. Now they’re building a company around it, too. In a crowded nondescript office in Sunnyvale, California, the 18-person team at #confluent is looking to share Kafka with companies like those users above and others like Goldman Sachs to change how companies think about their data. “This is almost like a tectonic shift in our foundations,” says CEO Jay Kreps. “These companies are increasingly built out of data.”
Investors see that shift and are lining up to get behind it. Already #backed by top firm Benchmark in its Series A, Confluent scored a $24 million Series B funding round announced on Wednesday led by Index Ventures. New board member Mike Volpi argues (here’s his blog post on the raise) that Confluent could prove the sexiest, and most lucrative, Web “plumbers” to date. “Kafka is perfect for today’s streaming applications,” Volpi tells Forbes. “Once we saw how ubiquitously Kafka was spreading, we knew this would become one of those iconic software companies.”
Sure—what any good investor would say. But Volpi was a part of a previous revolution at Cisco and is an investor in Elastic and Hortonworks. Confluent’s viral adoption so far was enough to excite Volpi to lead the large funding round just 8 months after its Series A brought the young company out of stealth.
Confluent’s a painfully young company—cofounder Neha Narkhede jokes that the funding’s use will be for a nice office for a change—but Narkhede stresses that as she, Kreps and cofounder Jun Rao built Kafka at LinkedIn, this is really a passion project years in the making.
The money will indeed go to a new office, says Kreps, but mostly will go into the product through new management tools to better use Kafka’s streams and eventually make money for Confluent in subscriptions. More analytics to monitor what’s happening in the streams like traditional data is tracked are also on the way.
But like other startups selling off a popular open-source technology such as the fast-rising Docker, Confluent will face a challenge in building a major business around its popular and free technology. The first step is to be popular at all. “Kafka is freeing up all your data locked away for years in silos,” Narkhede says. “So there’s plenty of excitement there.”
Featured Image credit to dbta.com