Upstash announced a $1.9 million seed round almost exactly two years ago, and an idea for building a serverless data platform for data-intensive applications built with Redis and Kafka. That doesn’t feel like that long ago, but a lot has changed in the data landscape since then.
Fast-forward to now and the company is boasting ARR of $1 million, a $10 million Series A led by a16z and a shiny new vector database. What’s changed in those two short years since the company launched is the growth of generative AI, which as you may have noticed is a pretty data-intensive application.
Company founder and CEO Enes Akar says that his idea to create a consumption-based pricing model, meaning you don’t pay until data starts moving through the system, has been a compelling one to developers, especially those using Kafka or Redis for data.
He says that the pricing model reduces the risk for creating applications. “Consumption pricing is our biggest differentiation in the market because it removes barriers for developers. As a developer, if I start a company I don’t have to pay until I get some real traffic and it becomes really popular,” Akar told TechCrunch.
The approach seems to be working with the company growing from 12,000 developers in March 2022 to 85,000 today. That kind of traction tends to attract attention, and Akar says Andreessen Horowitz actually approached him about funding, not a story you hear much these days in times of tightening purse strings from investors.
He got an unsolicited email last June from a16z offering a meeting, but Akar says he wasn’t looking for additional funding. In fact, he was more interested in staying independent while he built up the business, but when they offered to have a chat over Indian food in downtown Palo Alto, he decided to at least have a chat, because he joked, he really wanted to try that restaurant anyway.
Over dinner it came up that he didn’t have to start a board of directors to get this funding, and that got his attention because of his desire to remain independent as long as he could. “I felt like it was still early. I still had money. I also had some aggressive objectives like creating new products, but I told them I wasn’t interested,” he said.
By August, however, as he watched generative AI begin to really take hold, he had second thoughts. Akar, who is from Eskisehir, Turkey, was back home on vacation when he decided to call back to see if they were still interested in providing the funding. It turned out that they were and they began negotiating. The deal closed about a month later.
The new vector database, which plays a key role in helping find information in generative AI applications is a direct response to the big interest in this technology, and is specifically targeting AI developers. “So before Upstash Vector, we had seen many developers using Upstash to develop AI applications together with OpenAI APIs or Hugging Face APIs and they use Upstash Redis for caching and storing data. Now we’ve introduced Upstash Vector, which they can use just like Pine Cone or other vector databases,” he said.
The company, which currently has around 16 employees, with many of the employees still based in Turkey, recently hired Melek Pelen Esin, his longtime friend, whose background includes an MBA from MIT, a stint at McKinsey and time at Facebook in a sales management position, as COO. She’s running the business side of the house, while Akar runs engineering. They are both based in Silicon Valley.
They plan to hire another 15 or so employees in the next year, concentrating on professional support and customer success.