Originally published in Bootcamp. The first essay in a trilogy on design, AI, and self-knowledge and the most personal. It begins with a memory: a mother who was a psychoanalyst, a living room, and conversations that moved from the everyday into philosophy, mythology, and the nature of God before nightfall. When she passed way, so did that space of absolute listening. The essay traces what emerged from that absence, not an attempt to replicate the human, but a different question: can technology create a better mirror for our own consciousness? From that question came Amia, a conversational AI designed to guide users through their own patterns through Socratic dialogue. Pedro documents the design decisions behind it with unusual candor: why Amia opens with a curated deck of reflection cards instead of a blank chat, sacrificing naturalness for depth. Why the visual language is oneiric rather than utilitarian or futuristic. How the progressive tier system was designed to earn intimacy rather than presume it, calibrating depth as the AI builds context. And the honest admission of what doesn't work yet: the tension between rich context and token cost, between an AI that remembers everything and one that is simply fast. The essay closes with the strategic decision to build for Brazil first a country with some of the world's highest anxiety rates and some of its most unequal access to psychological care, as both a social urgency and a design laboratory.