Real-time communication intelligence.
What is said — synchronised, to the word, with what the voice and the face are doing at the same instant.
Runs in the browser · camera + mic required
AI has learned to understand language. But human communication was never just language. Presense synchronises the verbal stream — what is actually said, word by word — with the non-verbal signals the body emits at the same moment: prosody in the voice, micro-expressions on the face, physiological markers pulled from video. Only once those streams are aligned can an AI respond to the whole person rather than the transcript.
Live transcript, topic tracking, keyword detection — the semantic layer that gives every other signal its meaning.
Pitch, rhythm, energy, pauses, filler words. The voice carries emotional signal that the words alone never reveal.
52 facial blendshapes at 30fps, gaze and iris, genuine versus performed expression — the face that precedes the word.
Cardiovascular markers — stress, arousal, presence — read straight from video pixels. No wearable, no contact.
Most tools analyse these signals in isolation. Presense timestamps and fuses them — what was said at the exact moment the voice wavered or the face shifted. That alignment is what turns raw signal into something an AI can actually act on.
“At 1:42 — the speaker said confident, but pitch dropped 18%, vocal energy halved, and a micro-expression of doubt appeared 0.3s before the word. Likely masking uncertainty.”
The signals are the same everywhere. What changes is the interpretation — so each room gets its own surface, tuned to the conversation happening inside it. This deployment runs the demo. The rest are listed because they are what the platform is made of, not because they are open today.
A participant session, about fifteen minutes, camera and microphone on. It is the script the research runs on — a fixed sequence of prompts, the same for everyone, so that what changes between two people is the person and not the questions. It needs its own access code — a participant in the study is not a visitor to the demo, and the two are counted apart.
Optosense is a Finnish deep-tech company working at the intersection of computer vision, AI and human communication. We are a small, focused team, and Presense is our flagship product.
The gap between what people say and what they actually feel is still unsolved. We’re building the bridge. Investors, partners and early customers welcome.
Get in touch →