With artificial intelligence taking center stage in the enterprise world, focusing on AI principles and the hygiene of the data estate is a seminal challenge that has become a major priority for companies.

For AI to be successful, speed, scalability and simplicity are essential AI principles that play out behind the scenes, making the presence of contextual storage and the simplification of the data estate fundamental, according to Raj Verma (pictured), chief executive officer of SingleStore Inc.

“When we are talking speed, we are talking about single-digit-millisecond speed,” Verma said. “When you’re talking about scaling, complexity never scales, only simplicity does. The three tenets that you really have is speed, scale and simplicity. If those are the three tenets onto AI, then you would need an application development platform which has millisecond response time, has petabyte scale and has simplicity.”

Verma spoke with theCUBE Research analysts John Furrier and Dave Vellante at the “Supercloud 6: AI Innovators” event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed why speed, scalability and simplicity are important AI principles, as well as the role that SingleStore plays.

Storage and compute are major AI components

With data becoming the lifeblood of AI, storage and compute fit into the picture based on the need to build models with a large sample size for better decision-making. As a result, they are important AI components when it comes to integrating AI into business operations, according to Verma.

“If you break down AI, it was really about storage and compute,” he said. “Fundamentally, you have to put that as the basis on which AI is built, data and compute. There is a need for compute for AI to prosper. Everything else, as they say, is detail.”

Since not all data is equal, SingleStore prompts scalable operations irrespective of the data at hand, thanks to its advanced storage architecture. As a result, SingleStore ensures that hot and warm data is captured without breaking the bank and then marries it with historical data, Verma pointed out.

“There’s hot data, there’s warm data and there’s cool data,” he said. “The most contextual data is the hot and the warm data, and the historical context is in the cool data. We made certain architectural decisions at SingleStore, which provides us a three-tier storage architecture, which is in-memory, disk and then object store. You can manage your hot data in memory, sort of warm data in disk, and then as it cools, it goes to object store. You get hot and warm data at the best TCU, and that’s really why we called it SingleStore.”

Given that bridging the gap between transactions and analytics without having to use many databases is a burning issue that enterprises are grappling with, eradicating redundancies in terms of data assets is of the essence. One way of attaining this is by simplifying the data estate, according to Verma.

“The future of databases will be a database where you can have transactional and analytical system in one, hence SingleStore,” he said. “If you have 399 points of failure, you’re never going to be effective in AI. That simplicity, married with the scale that I was talking about, warm, hot and cold data, is the recipe for AI.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of the “Supercloud 6: AI Innovators” event:

Photo: SiliconANGLE