There is a new startup dedicated to commercializing Apache Flink, a distributed stream processing framework and dataflow architecture that emerged in Germany over a decade ago. Immerok, which has offices in Berlin and New York City, already offers a serverless Flink service on the cloud as part of its early access program.
Apache Flink has been a promising real-time processing framework since becoming a top-level project of the Apache Software Foundation in 2015. When Flink co-creators Stephan Ewen and Kostas Tzoumas founded his company DataArtisans in 2016, Flink would eventually become a commercial success, much like Databricks would eventually do with Apache Spark. It looked as if
In January 2019, Chinese web giant Alibaba, a large Flink user, reportedly acquired Data Artisans for €90 million (approximately $79 million at the then exchange rate). Alibaba offers a Flink service hosted on the public cloud and leverages the team to support its own in-house version of his Flink called Blink, but the idea of an independent commercial Flink provider is over. It looked like (data Artisans is still owned by Alibaba, although he changed his name to Ververica in January 2022).
While the open source Apache Flink project continues to evolve, other Flink backers wanted to give another try to independent commercial organizations backing real-time data processing frameworks. That finally happened in May when his Immerok was founded by a team of Flink committee members and PMC members.
Immerok yesterday announced a $17 million seed funding round backed by CUSP Capital, 468 Capital, Cortical VC and Essence VC, not to mention angel investors including Ewan. The company’s goal is to allow customers to run his Flink in his native way in the cloud without operational overhead.
“Our design goal is that everyone can leverage the full power of the Flink API to meet a variety of real-time business needs, from high-speed data pipelines and user-facing analytics to real-time ML/AI and transaction processing. Immerok co-founder Konstantin Knauf is also a member of the Flink PMC and one of Flink’s original co-creators.
The serverless Flink service on AWS is still in early access mode, but the company is already working with several customers as it aims to grow from 20 to 30 employees by the end of the year. One of them is his ING. Dutch bank with over $1 trillion in assets and standardizing his Flink for streaming applications.
“The next step in this evolution is to take a SQL-first approach and integrate more closely with the data lakes where machine learning models are trained,” Erik de Nooij, engineering lead at ING, presses. stated in the release. “This next step puts ING on track to put analytics capabilities into the hands of a wider audience for more data-driven decision-making.”
After leaving Ververica, Ewan is working with Immerok as an advisor and sees a lot of potential in the platform.
“The future of Apache Flink is thanks to Immerok,” Ewen said in a press release. “I have worked with the Immerok co-founders for years. I love working with.”
Users can request early access to Immerok Cloud here
Now that stream processing is finally starting to go mainstream, it’s the perfect time for Immerok. The company said Hadoop co-creator Doug Cutting once said he had a data flow architecture that outperformed other popular approaches such as Spark Streaming. It has the potential to position itself as an important part of the next stage in the evolution of
Immerok co-founder and CEO Holger Temme said in a press release: “Stream processing, and with it Apache Flink, is the foundation of these systems, and until now, organizations needed experts to build, operate, and scale these streaming applications—expensive and scarce experts.” Immerok makes it easier for organizations to build modern real-time systems by bridging the existing gap between stream processing and cloud-native applications.”
Related products:
Is real-time streaming finally in full swing?
Alibaba Acquires Apache Flink Backer Data Artisans
Flink Distro now available from data Artisans
