The growing use of AI agents throughout the enterprise is forcing a thorough reevaluation of the data layer.

This shift is driven by the need for millisecond responses that enable agents to make decisions, access data rapidly and integrate it fully into enterprise applications. Legacy batch-oriented architectures were not designed for this challenging environment, which has created a prime opportunity for real-time analytics, data warehousing and observability company ClickHouse Inc.

ClickHouse, which is available natively as a fully managed service on AWS Marketplace, is a real-time analytical database built to handle high concurrency and sub-second query performance in the AI era.

“AI needs very good context, very granular data,” said Tanya Bragin (pictured, right), vice president of product and marketing at ClickHouse. “Full fidelity data at scale is an absolute requirement. ClickHouse was actually built exactly for that customer-facing application at very high scale. The name ClickHouse is Clickstream Data Warehouse. Clickstream is just all of the users going to an intranet. That’s what it was built for.”

Bragin spoke with Christophe Bertrand, principal analyst at theCUBE Research, during an interview for the AWS Marketplace Series on theCUBE, SiliconANGLE Media’s livestreaming studio. She was joined by Sowmya Narayanan (left), director of product, billing and marketplaces at ClickHouse, and they discussed how the company is building a database to enable enterprise deployment of AI agents and applications at scale. (* Disclosure below.)

Natural language powers the data layer

ClickHouse provides a column-oriented, SQL-based database management system that focuses heavily on high throughput, low-latency queries, with efficient data compression and scaling across distributed clusters. Narayanan provided a demonstration of the ClickHouse platform during the interview. She illustrated the full user journey, from spinning up directly on AWS Marketplace to ingesting raw data and guiding the interface with natural language to receive a full analytical report in return.

“It’s gathering all of the data, and it’s assembling this complete answer,” Narayanan said. “It’s a structured report, it has visual elements and it has a detailed explanation of the lifecycle. It’s a full workflow from getting started with ClickHouse on the AWS Marketplace to provisioning the ClickHouse service, to ingestion, to analysis.”

The flow of the ClickHouse solution is also shaped by a database design expressly architected for the AI era. This can be seen in the use of a conversational interface within the platform to generate results, a key element in the transformation of the data layer, according to Bragin.

“The secret no one talks about is that for your AI initiatives to succeed, you absolutely have to get your data layer right,” she explained. “The introduction of conversational interfaces, for instance, to a data warehouse is completely changing assumptions around data platforms. More and more internal data teams are turning to a technology … to power a modern data warehouse that is conversational in nature or maybe fully autonomous where agents are running analysis and bringing insights to business leaders like myself, as opposed to me having to go and craft SQL.”

The company’s listing on AWS Marketplace allows users to bypass long procurement cycles that can often accompany SaaS opportunities. Customers can draw down against existing AWS committed spend, and a free trial is also available.

“The Marketplace listing is designed around how buyers actually want to work today,” Narayanan said. “Cloud-native customers live in AWS, and we want to meet them where they live. With ClickHouse, you can kick the tires in three clicks.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the AWS Marketplace Series:

(* Disclosure: TheCUBE is a paid media partner for the AWS Marketplace Series. Neither AWS, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE