Creativeapril - Fotolia
The new Swim DataFabric platform aims to help IT professionals categorize and make sense of large volumes of streaming data in real time.
The startup, based in San Jose, Calif., emerged from stealth in April 2018, with the promise of providing advanced machine learning and artificial intelligence capabilities to meet data processing and categorization challenges.
With the new Swim DataFabric, released Sept. 18, the vendor is looking to help make it easier for more users to analyze data. The Swim DataFabric platform integrates with Microsoft Azure cloud services including IoT suite and Data Lake Storage to classify and analyze data, as well as helps make predictions in real time.
The Swim DataFabric platform helps users get the most out of their real-time data with any distributed application including IoT and edge use cases, said Krishnan Subramanian, Rishidot Research chief research advisor.
"Gone are those days where REST is a reasonable interface for real-time data because of latency and scalability issues," Subramanian said. "This is where Swim's WARP protocol makes more sense and I think it is going to change how the distributed applications are developed as well as the user experience for these applications."
Why the Swim DataFabric is needed
Krishnan SubramanianChief research advisor, Rishidot Research
A big IT challenge today is that users are getting streams of data from assets that are essentially boundless, said Simon Crosby, CTO at Swim. "A huge focus in the product is on really making it extraordinarily simple for customers to plug in their data streams and to build the model for them, taking all the pain out of understanding what's in their data," Crosby said.
Swim's technology is being used by cities across the U.S. to help with road traffic management. The vendor has a partnership with Trafficware for a program that receives data from traffic sensors as part of a system that helps predict traffic flows.
The Swim DataFabric platform moves the vendor into a different space. The Swim DataFabric is focused on enabling customers that are Microsoft Azure cloud adopters to benefit from the Swim platform.
"It has an ability to translate any old data format from the edge into the CDM (Common Data Model) format which Microsoft uses for the ADLS (Azure Data Lake Storage) Gen2," Crosby said. "So, a Microsoft user can now just click on the Swim DataFabric, which will figure out what is in the data, then labels the data and deposits it into ADLS."
With the labelled data in the data lake, Crosby explained that the user can then use whatever additional data analysis tool they want, such as Microsoft's Power BI or Azure Databricks.
He noted that Swim also has a customer that has chosen to use Swim technology on Amazon Web Services, but he emphasized that the Swim DataFabric platform is mainly optimized for Azure, due to that platform's strong tooling and lifecycle management capabilities.
Swim DataFabric digital twin
One of the key capabilities that the Swim DataFabric provides is what is known as a digital twin model. The basic idea is that a data model is created that is a twin or a duplicate of something that exists in the real world.
"What we want is independent, concurrent, parallel processing of things, each of which is a digital twin of a real-world data source," Crosby explained.
The advantage of the digital twin approach is fast processing as well as the ability to correlate and understand the state of data. With the large volumes of data that can come from IoT and edge devices, Crosby emphasized that understanding the state of a device is increasingly valuable.
"Everything in Swim is about transforming data into streamed insights," Crosby said.