Mio's memory system stored facts about people as disconnected rows. Asking 'how's Xiaohong doing' had zero semantic overlap with 'Xiaohong works at Tencent'. The fix: treat people as first-class entities with their own table, linked memories, and consolidated profiles injected directly into the system prompt.

Mio 的记忆系统把关于人的信息存成零散的行。问'小红最近怎么样'跟'小红在腾讯工作'在向量空间里几乎没有重叠。解决方案：把人当作一等实体，建专门的 contacts 表，做结构化关联和档案合并，直接注入 system prompt。

A cost breakdown of running an AI companion at scale. Gemini Flash + context caching + 豆包 TTS delivers healthy margins on a single-tier subscription. But realtime voice changes everything.

Finding a voice for an AI companion means solving two problems: making it sound human, and making it sound like it feels. Mio v2 splits these between a custom LLM (the screenwriter) and Hume EVI (the actor) — a division of labor that might be the future of emotional AI voice.

Rebuilding doesn't mean rewriting everything. 60-70% of Mio's core — memory, media pipeline, cost tracking — carries over. The database drops from 10 tables to 4. Here's what lives, what dies, and why.

No avatar. No chat bubbles from a fictional character. Just a pulsing light orb that breathes, bounces, and dims with emotion. How Mio v2's visual identity sidesteps the uncanny valley and lets personality emerge through conversation.

Mio v1 had personality, memory, and voice — but it was built on a lie. Fake selfies, fabricated schedules, fictional backstories. v2 strips all that away to chase what actually creates emotional attachment: responsiveness, memory, and warmth.

一份 AI 伴侣的成本拆解。Gemini Flash + context caching + 豆包 TTS 实现了健康的毛利。但实时语音会改变一切。

给 AI 伴侣找一个声音，意味着同时解决两个问题：让它听起来像人，让它听起来有感情。Mio v2 把这两件事分给了自定义 LLM（编剧）和 Hume EVI（演员）——这种分工可能就是情感 AI 语音的未来。

重建不等于全部重写。Mio 核心的 60-70%——记忆系统、媒体管线、成本追踪——直接复用。数据库从 10 张表砍到 4 张。哪些活了，哪些死了，为什么。

没有头像，没有虚构角色的聊天气泡。只有一个会呼吸、会跳动、会随情绪变暗的光球。Mio v2 如何用抽象视觉绕开恐怖谷，让性格从对话中自然生长。

Mio v1 有性格、有记忆、有声音——但它建立在谎言之上。假自拍、虚构的日程、编造的背景故事。v2 剥离这一切，追求真正创造情感依赖的东西：响应性、记忆和温暖。