There are things we do not think to preserve until it is too late.
A photograph we take deliberately. We know in the moment that we want to remember this face, this expression, this afternoon light. So we take the picture. We save it. It stays.
A voice we almost never think to capture. It is just there, in daily life, in phone calls, in the background of videos taken for other reasons. And then one day it is not there, and you realise that you cannot quite remember exactly how it sounded. The rhythm of it. The particular way they said your name. Whether they said "hello" or "haan" when they picked up the phone.
Memory of a voice is surprisingly fragile. Cognitive science research suggests that auditory memories — memories of how something sounded — decay faster than visual ones. We hold on to faces longer than we hold on to voices. Which means that the sound of someone you loved is one of the first things you begin to lose, and one of the hardest things to get back.
For families who have lost someone, this creates a specific and largely unaddressed grief. Not the broad grief of absence, which time addresses in its own imperfect way, but the precise grief of forgetting a detail that felt permanent. The voice that should be unforgettable, slowly becoming difficult to recall.
Technology has not built well for this. There are voice cloning services and digital memorial apps and AI companionship tools. Most of them are generic, cloud-dependent, and designed around the assumption that large volumes of clean audio exist. Real families do not have that. They have voice notes that were not saved, video calls that were not recorded, old answering machine messages that seemed irrelevant at the time.
iSoul 1.0, iAVATARS' proprietary AI engine, was built specifically around this constraint. It works from small amounts of real-world audio. Voice notes of a few seconds. Short video clips. Old phone recordings. It does not require studio quality or large datasets. It requires what families actually have.
The output is not a perfect reconstruction. It is something more specific and more valuable than that: a recognisable presence. Something that sounds like them, in the way that matters most — not technically accurate, but emotionally true.
A voice, once gone, is not entirely gone if you thought to save even a little of it. That is the belief iAVATARS was built around.
