Betting on a startup is always a risky venture, but with Observe Inc., an application observability platform, Jeremy Burton (pictured), the company’s chief executive officer, seems to have hit the jackpot.

Observe is built on Snowflake Inc.’s data platform, allowing Observe to pool large amounts of data into a central data lake. This investment in a big data company has paid off, according to Burton.

Observe CEO Jeremy Burton discusses how the company’s observability platform is addressing query latency.

“We used to have wells in our back garden and generators in our basement … that’s where we were coming from back then, which is, let’s not look at where the market is today, where’s it going to be in 10 years’ time,” he said, explaining Observe’s origin. “We felt that customers are going to pay us for answering higher-level questions that they would have about their infrastructure or their applications or their business. If we could rely on someone else to do the hard work of the analytics and give them a few years to mature their platform, we figured we’d be in pretty good shape.”

Burton spoke with theCUBE’s Rebecca Knight and Dave Vellante at Data Cloud Summit, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how Observe’s relationship with Snowflake and how artificial intelligence impact observability. (* Disclosure below.)

An observability platform that synthesizes increasing data loads

Observe’s cloud platform analyzes telemetry, or diagnostics data, to detect issues in company applications, with a custom query language, OPAL, that can address the timing of different events.

“Data volumes are growing 40% a year, particularly unstructured data. So, scale is still very, very important,” Burton said, emphasizing the value of resiliency in the data environment and, most importantly, low latency. “These features are not like the sexy features that you would maybe show in a demo, but they’re features that you’re really not going to get into a large enterprise unless you can prove you can do.”

One of Observe’s biggest partners is Capitol One, and they recently finished a series B, with Snowflake and Sutter Hill Health making major investments. Last year, the company had an estimated net revenue retention of 174% and now does 120 million Snowflake queries a day.

“When the data comes into Snowflake, we transform that data …  if it’s data coming in from cloud trails, you can look at EC2 instances or S3 buckets, or if you’ve got an application, you can look at customers and sessions and transactions,” he said. “That transformation of data and then the point-and-click interface to troubleshoot and analyze that data, that’s our value add.”

Tackling large amounts of data is, of course, crucial for the era of large language models. Although Observe uses a chat bot to help users learn OPAL, Burton has doubts about the AI hype that is currently sweeping the industry.

“The current technology, large language models, they don’t have an ability to understand or to reason. I think as a productivity improvement, they’re going to be great. Nobody ever read the documentation on anything we ever wrote. Now people will be able to get it through a chatbot,” Bruton said. “What I question is the underlying model and the hallucination and then the amount of money it takes to train. Are there use cases where people will pay such that you can actually make a business out of it? That’s what I’m yet to be convinced of.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of Data Cloud Summit: 

(* Disclosure: Snowflake Inc. and Observe Inc. sponsored this segment of theCUBE. Neither Snowflake, Observe nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE