Nimble Way raises $47M to give AI agents access to real-time web data

Believe it oregon not, web hunt is inactive thriving arsenic an industry. As businesses put successful utilizing AI agents to marque the astir of their data, there’s request for tools that not lone scrape the web to pass what those AI bots do, but besides instrumentality those results successful a mode that’s easier to usage with modern information tools.

That’s the committedness down web hunt startup Nimble Way, which precocious raised a $47 cardinal Series B round, led by Norwest. The New York company’s level employs AI agents to hunt the web successful existent time, verify, and validate the results, and past operation the accusation into neat tables that tin past beryllium queried similar a database.

That past portion is important here. LLMs and AI agents are large for searching the web, connecting results from a assortment of sources, and analyzing them, but they often instrumentality the results successful plain text, which tin beryllium hard to enactment with astatine an endeavor level. And that’s earlier you origin successful hallucinations, the hazard of the cause misunderstanding your instructions, oregon the usage of unreliable sources.

By validating and structuring results into tables, Nimble Way lets companies usage web information arsenic if it were already portion of their existing databases. The startup besides integrates with endeavor information warehouses and information lakes — ample centralized repositories wherever businesses store and analyse information — offered by the likes of Databricks and Snowflake. That means its AI agents tin plug into a business’s trove of data, utilizing it to physique context, and signifier however hunt results are structured and returned.

In effect, this lets enterprises person live, structured web information arsenic portion of their existing information environments, Nimble Way CEO and co-founder Uri Knorovich (pictured above, middle) told TechCrunch.

Such integrations besides let Nimble Way’s bundle to retrieve constraints — specified arsenic however you privation the hunt to beryllium performed, oregon which information sources to tap. This is peculiarly utile for applications specified arsenic rival analysis, pricing research, know-your-customer (KYC) processes, marque monitoring, heavy research, and fiscal analysis. (Knorovich noted that Nimble Way works to guarantee each lawsuit information remains wrong customers’ information infrastructure to comply with information retention and information policies.)

To that end, the startup has partnered with Databricks, Snowflake, AWS and Microsoft to assistance streamline endeavor deployments that necessitate entree to interior information sources. (Databricks besides participated successful this Series B.)

“Models tin bash a batch of things, but astir accumulation AI fails aren’t due to the fact that the models are not bully capable — it’s due to the fact that of a information failure,” Knorovich said. “What we’re seeing contiguous is that enterprises don’t request much AI; they request AI with good, reliable web hunt […] If you nail it down, if you tin take what your cause tin hunt and cannot search, this is the tipping constituent for enterprises to say, ‘hey we tin really spot AI. We tin really enactment AI to enactment successful much usage cases’.”

Knorovich says the quality to hunt the web successful existent clip astatine scale, and validate and operation hunt results, is what sets Nimble Way isolated from different information brokers already successful the space.

The startup presently has much than 100 customers, with the bulk of its gross coming from ample enterprises, Fortune 500 companies, and adjacent immoderate Fortune 10 companies, including large retailers, hedge funds, banks, and user packaged goods companies, arsenic good arsenic immoderate AI-native startups.

“Nimble is tackling a occupation that has existed for years without a due solution and is present becoming of captious urgency,” Assaf Harel, spouse astatine Norwest, said successful a statement. “Trusted unrecorded web information is progressively becoming a prerequisite for AI agents performing captious concern decisions.”

The Series B besides saw information from returning investors Target Global, Square Peg, Hetz Ventures, Slow Ventures, R-Squared Ventures, J-Ventures, and InvestInData. Proceeds from the circular volition beryllium utilized to grow R&D successful multi-agent web hunt and a governed information furniture that processes and validates hunt results.

Nimble Way has present raised a full of $75 million.