WisdomAI burst out of stealth mode, announcing a $23 million seed round led by Coatue Ventures with contributions from Madrona, GTM Capital, The Anthology Fund, U First Capital, Latitude Capital, and over 30 angel investors. The San Francisco-based startup was founded by Soham Mazumdar and a team of data engineers, who are all former colleagues at Rubrik and bill themselves as delivering the first “Agentic Data Insights Platform” built on what it calls a “Knowledge Fabric.” In WisdomAI’s vision, this fabric is a dynamic, interconnected map of a company’s data and domain context continuously refined by human experts that serves as the foundation for AI reasoning. In short, instead of rigid dashboards or scheduled reports, WisdomAI uses intelligent software agents to actively surface business answers from the data.
This approach is meant to address a common frustration in large enterprises. Despite massive investments in data warehouses and analytics tools, many companies find that traditional dashboards “fall short” of true self-service, leaving users locked into pre-defined reports. As ex-Tableau CEO Mark Nelson puts it, business users are often “limited to the insights that have been pre-defined by the analyst who created the dashboard,” and any new question can stall a project for days or weeks.
Inside the Knowledge Fabric
WisdomAI’s founders argue that the remedy is a layer of AI that truly understands the business’s own data language. The Knowledge Fabric they describe encodes terminology, key performance indicators, and data relationships so that answers can be generated on the fly. For example, the company explains that its fabric is not a static model but a “living reflection of how your business runs,” continuously shaped by analysts and domain experts. In practice this means a marketing or sales manager can simply ask a question in plain English—say, about quarterly forecasts or campaign performance—and the system will translate that into database queries across its data silos, returning a precise answer in seconds.


