It may come as a surprise, but the web search industry continues to flourish. As companies increasingly adopt AI agents to leverage their data, there is a growing need for tools that not only gather web information to guide these AI systems but also present the findings in formats compatible with contemporary data platforms. This is the core mission of Nimble, a web search startup that has just secured $47 million in a Series B funding round led by Norwest.

Based in New York, Nimble’s platform uses AI agents to perform real-time web searches, verify and validate the information retrieved, and then organize it into structured tables. These tables can subsequently be queried much like a traditional database. This final step is particularly significant. While large language models and AI agents excel at scouring the web, synthesizing information from diverse sources, and performing analysis, they typically deliver results in plain text. This format can be cumbersome for enterprise-level applications, not to mention the additional challenges of AI hallucinations, misinterpreted instructions, or reliance on dubious sources.

By validating information and structuring it into tables, Nimble enables businesses to utilize web data as seamlessly as if it were already integrated into their internal databases. The startup also offers integrations with major enterprise data warehouses and data lakes, such as those provided by Databricks and Snowflake. These are large, centralized systems where organizations store and analyze their data.

This connectivity allows Nimble’s AI agents to tap into a company’s proprietary data, using it to add context and influence how search results are formatted and delivered. Furthermore, these integrations enable the software to retain specific constraints—like preferred search methodologies or approved data sources—tailoring the process to each business.

Such capabilities prove especially valuable for tasks like competitor analysis, pricing research, know-your-customer (KYC) procedures, brand monitoring, in-depth research, and financial analysis. (Knorovich emphasized that Nimble ensures all customer data stays within the client’s own data infrastructure to adhere to data retention and security policies.)

To facilitate enterprise deployments that require access to internal data, the startup has established partnerships with Databricks, Snowflake, AWS, and Microsoft. (Databricks also contributed to this Series B round.)

“Models are capable of many things, but most failures in production AI aren’t due to inadequate models—they stem from data failures,” Knorovich stated. “What we’re observing now is that enterprises don’t necessarily need more AI; they need AI supported by reliable, high-quality web search […] If you can precisely control what your agent is permitted to search, that becomes the tipping point. It allows enterprises to confidently say, ‘We can actually trust AI. We can deploy it across more use cases.’”

According to Knorovich, Nimble’s distinction lies in its ability to perform large-scale, real-time web searches while also validating and structuring the results—a contrast to other data brokers in the market. The company currently serves over 100 clients, with the bulk of its revenue derived from large enterprises, including Fortune 500 and even some Fortune 10 companies. Its customer base spans major retailers, hedge funds, banks, consumer packaged goods firms, and several AI-native startups.

“Nimble is addressing a long-standing problem that lacked a proper solution and has now reached a critical level of urgency,” said Assaf Harel, a partner at Norwest, in a statement. “Trusted, live web data is fast becoming a prerequisite for AI agents making crucial business decisions.”

The Series B round also included participation from returning investors Target Global, Square Peg, Hetz Ventures, Slow Ventures, R-Squared Ventures, J-Ventures, and InvestInData. The newly acquired funds will be directed toward expanding research and development in multi-agent web search and a governed data layer responsible for processing and validating search results. To date, Nimble has raised a total of $75 million.