Reference
Glossary
Key terms in AI emotional wellness — what they mean, why they matter, and how Cherizh uses them.
Last reviewed:
- AI Companion
- An artificial intelligence application designed to provide ongoing conversational support, companionship, and emotional presence. Unlike AI therapy apps that deliver structured clinical techniques, AI companions focus on relationship-building, consistent availability, and making users feel known and remembered. Examples include Cherizh, Replika, and Pi.Learn more →
- AI Emotional Wellness
- The use of artificial intelligence technology to support everyday emotional health — not clinical treatment, but the daily emotional check-ins, pattern recognition, and companionship that help people understand and manage their emotional lives. AI emotional wellness fills the gap between therapy sessions, covers the 2 AM moments, and provides support without appointment waitlists.Learn more →
- AI Therapy App
- An AI application that delivers structured clinical techniques such as Cognitive Behavioral Therapy (CBT) or Dialectical Behavior Therapy (DBT) through guided exercises and assessments. AI therapy apps like Woebot and Wysa focus on symptom management rather than companionship. They are not replacements for licensed human therapists.Learn more →
- AES-256 Encryption
- Advanced Encryption Standard with 256-bit key length — the same encryption standard used by banks, government agencies, and military organizations to protect classified information. Cherizh uses AES-256 encryption for all user data in transit and at rest, meaning your conversations are protected at the highest commercially available standard.Learn more →
- Contextual Memory
- An AI system's ability to remember not just what you said, but the emotional context, relationships mentioned, mood at the time, and how that conversation connects to previous ones. Contextual memory is what makes the difference between an AI that recalls facts and one that genuinely understands your story. Cherizh's 4-layer memory architecture is built on contextual memory.Learn more →
- Crisis Detection
- The ability of an AI emotional wellness app to recognize when a user's emotional state may indicate crisis-level distress — not just a bad day, but patterns suggesting escalating risk. Effective crisis detection includes pattern-based early warning, warm handoff to professional crisis resources (like the 988 Lifeline), and clear scope boundaries about what the app can and cannot do.Learn more →
- Emotional Pattern Recognition
- The AI capability to identify recurring emotional trends over weeks and months — such as anxiety that peaks on Sunday nights, mood shifts tied to specific relationships, or gradual improvements the user can't see themselves. This is the third layer of Cherizh's memory architecture and one of the most valuable features for emotional self-awareness.
- Identity Memory
- The deepest layer of Cherizh's 4-layer memory architecture. Identity memory captures who you are — your values, your communication style, what matters to you, and how your sense of self has evolved over time. This layer enables the AI to understand you as a whole person rather than a series of disconnected conversations.
- Living Memory
- A persistent, contextual AI memory system that grows and evolves with the user over time. Unlike traditional chatbots that reset with each session, living memory creates continuity — the feeling of being known. Cherizh's living memory architecture encompasses conversation recall, relationship mapping, emotional pattern recognition, and identity memory. The term reflects the fact that this memory is not static storage but a dynamic understanding that deepens.Learn more →
- Mood Tracking
- The systematic recording and analysis of a user's emotional states over time. In AI emotional wellness apps, mood tracking goes beyond simple "rate your day" inputs — it includes analysis of emotional language in conversations, identification of triggers and patterns, and visualization of emotional trends. Cherizh's mood tracking uses pattern recognition to surface insights the user may not notice themselves.
- Relationship Mapping
- The second layer of Cherizh's memory architecture. Relationship mapping tracks the people in your life — how you talk about them, the emotional weight of each relationship, recurring dynamics, and changes over time. This enables conversations like "You've mentioned Sarah three times this week, each time with tension in your words — want to talk about what's happening there?"
- Sit With Me™
- A proprietary feature of Cherizh that provides silent companionship — presence without words. Designed for moments when you don't need advice, solutions, or conversation, just the knowledge that someone is there. Sit With Me recognizes that sometimes the most powerful form of support is simply showing up and sitting with someone in their experience without trying to fix it.
- The 2 AM Problem
- The gap between needing emotional support and being able to access it — particularly acute at night, on weekends, and during holidays when therapists are unavailable, crisis lines may be overwhelmed, and friends are asleep. Coined by Cherizh founder Kelly Kuo to describe the specific kind of isolation that exists when you need someone and the system isn't there. AI emotional wellness apps exist primarily to address this gap.Learn more →
- Therapeutic Alliance
- The relationship and trust between a person and their support system — whether human therapist or AI companion. Research from Stanford's Human-Centered AI Institute shows that the key factor in therapeutic alliance isn't technique or intelligence — it's continuity, the feeling that someone knows your story. Living memory AI is built on this insight.
- Voice Journaling
- A feature that allows users to speak their thoughts aloud rather than typing. The AI listens, processes the content and emotional tone, and helps the user understand patterns in what they're expressing. Voice journaling lowers the barrier to emotional processing — particularly valuable when typing feels too effortful or when speaking reveals emotions that writing might filter.
- Warm Handoff
- The process of transitioning a user from an AI companion to professional crisis resources in a way that feels supportive rather than abrupt. A warm handoff includes context ("What you're describing sounds like it might need more than I can offer right now"), guidance ("Here's who can help, and here's why they're good at this"), and post-crisis continuity (remembering and not requiring the user to re-explain when they return).
- Witnessing Model
- Cherizh's core design philosophy. Unlike clinical apps that deliver therapeutic exercises (intervention model), the witnessing model holds that most people need to feel seen before they can heal. Cherizh validates emotions first, problem-solves second. The approach is rooted in the belief that acknowledgment — "I see how heavy this is for you" — is itself therapeutic.Learn more →