About us

We're a team of engineers with 20 years of experience in machine learning and BI, and a love for well-crafted technologically-advanced products. We're also repeat entrepreneurs and proud YC alums (W17).

Now that the world is reaching peak AI hype, it's time to reinvent the enterprise with AI systems that can actually deliver on their promises.

The BI industry has been pitching self-served BI for more than three decades, yet everywhere we look, non-technical people still rely heavily on data teams to answer their questions. While natural language in a chat interface seems like the best paradigm to achieve this, LLM-based products fall short because of hallucinations, a problem that is intrinsic to their inner workings.

Our vision is that BI is not strictly a language problem, and thus requires a different approach than slapping an LLM + RAG. We are on a mission to design the most performant AI specifically for business intelligence, with the hard constraint that it should NEVER hallucinate: zero hallucinations is the only way to appropriately serve the needs of non-technical people who by definition can't verify answers and need to fully trust the system.

After years of research, we developed a proprietary agentic architecture that will finally make it possible for enterprises to trust and deploy AI-powered BI systems to thousands of employees, and eliminate a huge part of BI tickets backlogs.