Turning voice into a vital sign - Canary Speech

What if your voice could reveal what a blood test cannot?

Our latest episode features Henry O'Connell, Founder and CEO of Canary Speech, alongside the Jeff Adams, a speech technology pioneer who led the development of Dragon by Nuance, wrote the code that became Siri, and helped build the Amazon Echo.

Canary Speech has developed groundbreaking technology that treats voice as a biomarker. Unlike traditional NLP that focuses on what words are spoken, their platform analyzes how they're spoken, capturing 12.5 million data elements in just 40 seconds of speech and analyzing 2,590 distinct speech elements every 10 milliseconds. The technology works ambiently during natural conversation, returning actionable results while the patient is still in the room.

Their platform currently addresses anxiety, depression, fatigue, and wellness, with expansion into PTSD, postpartum depression, ADHD, autism, and beyond. Through partnerships with Harvard, Beth Israel, NYU, Mayo Clinic, Cleveland Clinic, and the NIH, they're building the clinical foundation for a new category of diagnostics. The technology has already been integrated into Microsoft's Dragon Copilot, bringing voice biomarker analysis directly into clinical workflows.

Henry shares a powerful analogy: just as a parent can gauge a child's wellbeing from their tone regardless of words spoken, Canary Speech provides clinicians with objective data to validate their intuition, serving as a "force multiplier" for medicine.

This episode is a must-listen for anyone interested in voice technology, AI-driven diagnostics, or novel approaches to mental health screening and early disease detection.