WisdomAI, the new AI data analytics startup from Rubrik co-founder Soham Mazumdar, has landed a fresh $50 million Series A led by Kleiner Perkins with participation from new investor NVentures (Nvidia’s venture capital arm).
This round comes roughly six months after the startup announced a seed round of $23 million led by Coatue.
WisdomAI offers AI-driven data analytics that can answer business questions from structured, unstructured, and even “dirty” data, meaning data that hasn’t been cleaned of typos or errors. A business user simply asks questions in natural language, such as, “How many customers do I have in my pipeline and what’s preventing them from closing this quarter?”
But the company is using a clever method to eliminate the LLM hallucination problem. WisdomAI is not using LLMs to write answers to questions. Instead, LLMs are only used to write the query — the part that will go out to a data warehouse to retrieve data. So if the LLM hallucinates, it will simply write an ineffective query rather than inventing false answers.
WisdomAI has written its own logic that it calls the “enterprise context layer,” which studies the customer data to understand it. All of the startup’s co-founders worked with Mazumdar at data security company Rubrik, giving them deep experience with enterprise storage warehouses. (Mazumdar left Rubrik in 2023.)
Since formally launching in late 2024, WisdomAI has grown from two enterprise customers to around 40 enterprise customers, CEO Mazumdar says. WisdomAI counts such companies like Descope, ConocoPhillips, Cisco, and Patreon as customers.
Mazumdar said the young company is also growing usage within its enterprise customers. Some customers have doubled usage within two months. Another customer started with 10 seats and expanded to 450, which is almost everyone in the company, he said.
In the last six months, WisdomAI has also added an agentic feature that will alert users in real time to important changes in situations they are monitoring.
“I have created an agent which is watching our product usage metrics, our ticket information,” Mazumdar says, adding that it took him about five minutes to create it. But rather than sending him a daily or, even, hourly report on usage and helpdesk tickets, it alerts him “when something interesting happens,” he describes.
“I think that’s the magic with analytics. It’s always been a static report, but we are making it dynamic. We are making it proactive,” he said.