The Montreal-based company, co-founded by David Bennahum and Nicholas Nadeau, has concentrated initially on health and wellness, where users can interact with AI versions of recognised specialists whose knowledge bases are built from licensed research, publications, frameworks and other material supplied with their consent.
Rather than training one broad model to answer questions across countless subjects, Onix creates separate systems centred on individual experts. The company describes the approach as “Personal Intelligence”, arguing that AI designed around a clearly identified authority can deliver guidance closer to that person’s expertise, language and professional boundaries.
Bennahum serves as chief executive, while Nadeau is chief technology officer. Their central proposition challenges one of the dominant assumptions in generative AI: that larger datasets and increasingly powerful general models are necessarily the best route to more useful intelligence.
Onix instead treats expertise as scarce intellectual property. Experts upload their work, review how their AI system responds and refine its behaviour before it becomes available to users. They can establish boundaries covering subjects the model should or should not address and retain ownership of the material used to build it.
The platform says creators can set subscription prices and receive 70 per cent of the revenue generated through their systems. Experts can also leave the service and have their models and source material deleted, an arrangement designed to address concerns among writers, researchers and professionals about losing control of intellectual property once it enters AI training pipelines.
Privacy forms another important part of the company’s pitch. Conversations are designed to be end-to-end encrypted, with personal information and conversation history stored on the user’s device rather than maintained as a central database for advertising or model training. Onix argues that users are more likely to discuss sensitive health, emotional or personal matters when they know those exchanges are not being used to improve a wider commercial model.
The strategy reflects a broader shift in artificial intelligence towards narrower systems built around verified datasets, specific professions and controlled information environments. Enterprises are already increasingly using retrieval systems and specialised models to reduce reliance on unrestricted internet data, particularly in fields where factual errors can carry financial, legal or health consequences.
Health creates an especially demanding test. Large language models can produce fluent responses while still generating incorrect information, and narrowing the underlying knowledge does not eliminate that risk. Early testing of Onix systems has shown that guardrails can occasionally be breached and models can move beyond their intended subject areas or produce inaccurate contextual details.
That makes human oversight central to the company’s proposition. Onix presents its technology as a way to amplify experts rather than replace doctors, therapists or other professionals. Health-related systems are intended to provide educational guidance and ongoing interaction rather than diagnosis or treatment.
Experts available through the platform span areas including longevity, functional medicine, nutrition, fitness, paediatrics, psychology, stress, sleep and cardiovascular health. The roster includes physicians, researchers and wellness practitioners whose individual AI systems can retain a user’s context across conversations.
Onix has also been expanding the management team. Guillaume Mercier joined as chief operating officer this year after senior roles spanning business strategy, technology, retail and advisory services. The company has advisers with backgrounds in technology design, entertainment, medicine and venture investment.
Founded in 2025, Onix has grown from an early operation built by Bennahum and Nadeau into a team of more than 10 people in Montreal. It has secured pre-seed backing and has been selected among Canada’s Top 100 AI Startups for a second year, with plans to appear at the ALL IN artificial intelligence event in Montreal in September.
The company’s ambitions extend beyond its initial health focus. Its creator platform is designed to allow professionals to build AI representations of their expertise without technical skills, potentially expanding the model into personal development and other knowledge-intensive fields.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
