In a crowded room, isolait finds the one voice you came for — and quiets everything else. Real-time. On-device. It never leaves your phone.
Every feature runs locally on Apple's Neural Engine — there's no server in the loop, no upload, and no delay.
Tap the mic to sample any voice nearby. Isolait captures a unique acoustic fingerprint in seconds, then locks onto that speaker — filtering out background noise and every other conversation in the room.
Point your phone at the person you want to hear. Isolait blends your device's heading with real-time audio processing to bias the mic toward your target, even in multi-speaker environments.
Watch your audio environment come alive with a dual-channel spectrum analyzer — see exactly what's coming in, and exactly what isolait is sending back out.
Isolait runs entirely on-device using Apple's Neural Engine. No audio is ever uploaded, stored, or shared — what you hear stays with you, full stop.
Most "AI" voice tools mean a round trip to a server. Isolait's neural pipeline runs entirely on your iPhone's Neural Engine — your audio is processed and discarded locally, every time.
v1.1 is live. Everything below builds on the same on-device engine — no architecture rewrite, just new capability layered onto the foundation that already works.
CoreML speaker embedding replaces the rule-based filter from v1.0 — locking onto a voice fingerprint instead of a frequency range. Lock screen controls and AirPods routing round out daily usability.
Real-time transcription and multi-speaker diarization turn isolait into a meeting companion — with named speakers that persist across sessions, no re-enrollment required. The same Voice Lock pipeline also unlocks a hearing-clarity mode: headphone playback of a locked speaker's voice, boosted for noisy environments.
Real-time translation built on a single isolated voice, rather than a noisy room, is dramatically more accurate. Longer-term, the same on-device engine becomes licensable infrastructure — a Dimalta AI Audio SDK for any product that needs to hear one voice clearly.
Free to download. Works the moment you open it — no account, no setup, no audio ever leaving your phone.