Alexandre LeBrun

From Paris to Silicon Valley: A Career Built on Conversational AI

Alexandre LeBrun’s journey in artificial intelligence began in Paris, where he studied at École Polytechnique before moving to the United States to build companies that shaped how humans interact with machines. His first major venture, VirtuOz in 2002, pioneered intelligent virtual agents—AI systems designed to automate customer interactions for enterprises like AT&T and Symantec. Unlike many early chatbots that relied on rigid scripts, VirtuOz incorporated proprietary natural language processing and machine learning to deliver personalized responses, a concept that would later define the industry. After bootstrapping the company in France, LeBrun moved it to San Francisco, securing funding and achieving profitability before Nuance Communications acquired VirtuOz in 2012.

LeBrun’s next project, Wit.ai, took his work in conversational AI further by creating a platform that allowed developers to integrate voice interfaces into applications without needing deep expertise in AI. Founded in 2013, Wit.ai gained traction for its ability to bridge the gap between raw speech input and actionable commands. The company’s approach caught Facebook’s attention, leading to its acquisition in 2015. LeBrun joined Meta’s AI research team under Yann LeCun, where he contributed to projects like Facebook’s M assistant, an AI that blended automation with human-like reasoning. His time at Meta reinforced his belief that while generative models excel at tasks like summarization and coding, they often rely on shortcuts that limit their understanding of the real world.

Nabla: Bringing AI to the Frontlines of Healthcare

In 2019, LeBrun co-founded Nabla, a healthcare AI startup focused on assisting clinicians rather than replacing them. The company’s AI assistant integrates with electronic health records to streamline documentation, reduce administrative burdens, and improve patient interactions. Today, Nabla supports over 85,000 physicians across various specialties, reflecting LeBrun’s commitment to building AI that addresses tangible, high-stakes challenges. His leadership at Nabla highlights a shift from consumer-facing AI to mission-driven applications, where reliability and precision are paramount.

LeBrun’s work in healthcare also underscores his broader philosophy about AI’s role in society. He has consistently emphasized the need for systems that go beyond predictive accuracy to achieve genuine understanding—a theme he revisits in his recent role as CEO of AMI Labs. His perspective is shaped by decades of observing AI’s evolution, from early virtual agents to today’s large language models, each revealing both the promise and limitations of current approaches.

AMI Labs: A New Chapter in Foundational AI

In late 2025, LeBrun took on a new challenge as CEO of AMI Labs, a company co-founded with Yann LeCun and others to develop advanced machine intelligence. AMI secured a $1.03 billion seed round, signaling strong investor confidence in its mission to build AI systems capable of comprehending the real world. LeBrun’s announcement highlighted his skepticism toward generative architectures as a path to true understanding, arguing that they often rely on tokenized, low-dimensional shortcuts. Instead, AMI aims to explore world models—systems that learn abstract representations of reality to predict the consequences of actions in high-dimensional, noisy environments.

This venture reflects LeBrun’s long-standing interest in the philosophical underpinnings of AI, as seen in his personal writing and public reflections. Whether through his LinkedIn posts or essays on platforms like Medium, he has consistently questioned the assumptions behind today’s most popular AI techniques. His career—spanning startups, acquisitions, and research—demonstrates a rare blend of technical innovation and critical thinking, positioning him at the forefront of AI’s next evolutionary leap.