Paris-based startup Clarifeye has raised €4 million in a pre-seed round led by EQT Ventures to tackle one of the biggest headaches in enterprise AI: most of it doesn’t work. Founded by a team of seasoned leaders from data science giant Dataiku, Clarifeye is betting that the key to unlocking generative AI for complex industries isn’t just feeding it more documents, but teaching it how to think.
The core problem, as many companies are discovering, is that today’s large language models are brilliant generalists but terrible specialists. They can write an email or summarize a report, but they can’t replicate the nuanced, instinct-driven reasoning of a 20-year veteran engineer or a top regulatory lawyer. This "tacit knowledge", the stuff that’s in people’s heads, not in a manual, is where most of the value lies. Its scarcity creates bottlenecks and, according to a recent MIT study cited by the company, is a key reason why 95 percent of corporate GenAI pilots are failing.
Current solutions like Retrieval-Augmented Generation (RAG) try to solve this by pointing an LLM at a pile of internal documents, but that often just results in a slightly more informed but still shallow chatbot. It can find the what, but not the why or the how.
Beyond RAG
Clarifeye is taking a different approach. Instead of just connecting AI to data, it’s building a platform to connect AI to experts. The system creates a “GenAI-ready data layer” where subject-matter experts can collaboratively map out their decision-making patterns, reasoning strategies, and the invisible connections they make between disparate pieces of information. This transforms informal, experience-based knowledge into a structured, operational rulebook that an AI agent can use.
The goal is to create AI agents that don’t just retrieve information but can reason through complex problems with the depth and accuracy of a company’s most trusted specialists. For industries like life sciences, manufacturing, or law, where a single decision can have massive consequences, this promises a level of reliability and traceability that generic AI can’t touch.
“Most GenAI efforts fail because companies are forced to choose between generalist AI, which is fast but shallow, or vertical AI, which is narrow but rigid,” said Mathieu Grisolia, CEO of Clarifeye, in a statement. “Clarifeye takes a different path: we capture what is in your experts’ minds, turning it into an always-on AI that thinks like them.”
The Clarifeye funding will be used to build out its engineering team and further develop the platform. The round also saw participation from Drysdale Ventures and prominent angels including Datadog CEO Olivier Pomel.
For EQT Ventures, the investment is a bet on bridging the gap between AI hype and actual business impact. “By turning tacit knowledge into collective intelligence, Clarifeye helps enterprises unlock not just the potential of AI, but the full potential of their people,” said Gustav von Sydow, Partner at EQT Ventures. It’s an ambitious plan, but if Clarifeye can successfully bottle the lightning of human expertise, it could redefine what’s possible for AI in the enterprise.
