I built Ontor to measure nervous system state from your voice during real calls
I've spent my career pulling useful signal out of noisy data: geophysics, factory vision, and radiology ML. Over the past few months I've applied the same kind of work to voice.
Ontor shows how stress, energy, confidence, and breathing shift during calls and presentations, relative to your own baseline. You can see the read during a session, review what changed afterward, and try a short reset before the next one.
There's an important limit to the claim. Ontor doesn't diagnose a condition or label an emotion. The individual voice markers come from published speech research, but the weights that combine them still need calibration against ground truth. I trust the within-person change more than any absolute score.
I built the signal processing, models, and clients myself. I'm now testing it with people who care about how they show up: biohackers, coaches, and anyone whose performance happens in conversation.
I'd value a blunt read from other builders. Where would this be useful, and where would it feel like noise?
