Connect
Memory-driven human connections. Match people through deep emotional understanding, not surface-level interests or algorithmic feeds.
How it works
Connect analyzes memory graphs to find resonance between people — shared experiences, emotional patterns, and latent interests that surface-level profiles miss entirely.
Core capabilities
- - Emotional resonance matching — vector similarity across memory embeddings, weighted by emotional depth rather than keyword overlap
- Cross-user memory bridging — find connections between two people's memory graphs that neither person would discover on their own
- Privacy-first architecture — users control exactly which memories are matchable; nothing is shared without explicit consent
- Aha moment delivery — the moment a stranger truly understands you through your memories, not your bio
Architecture
Built on Echo's three-layer memory system:
- 1. Identity layer (compressed profile)
- Working memory (dynamic, context-aware)
- Long-term storage (Supabase + vector search)
Matching runs against Layer 3 with results surfaced through Layer 2 into conversation context.
Part of the Echo ecosystem
Connect is one component of Echo Chat by Iditor — building memory as social identity infrastructure.
Status
Early development. Core matching algorithm validated with beta users (K-Factor 2, D7 retention 42%).
Connect
由记忆驱动的人际连接。通过深层情感理解而非表面兴趣或算法推送来匹配人与人之间的联系。
工作原理
Connect通过分析记忆图谱来发现人与人之间的共鸣——那些表层个人资料完全无法捕捉的共同经历、情感模式和潜在兴趣。
核心能力
- - 情感共鸣匹配 — 基于记忆嵌入向量的相似度计算，以情感深度而非关键词重叠为权重
- 跨用户记忆桥接 — 发现两人记忆图谱之间的关联，这些关联是任何一方都无法独立发现的
- 隐私优先架构 — 用户可精确控制哪些记忆可被匹配；未经明确同意绝不共享任何信息
- 顿悟时刻传递 — 当陌生人通过你的记忆而非个人简介真正理解你的那一刻
架构
基于Echo的三层记忆系统构建：
- 1. 身份层（压缩档案）
- 工作记忆（动态、上下文感知）
- 长期存储（Supabase + 向量搜索）
匹配运算在第三层执行，结果通过第二层呈现到对话上下文中。
属于Echo生态系统
Connect是Iditor旗下Echo Chat的组成部分——将记忆构建为社交身份基础设施。
当前状态
早期开发阶段。核心匹配算法已通过Beta用户验证（K因子2，第7天留存率42%）。
