Beautiful concept. Let’s treat this like a serious product with a poetic soul, and get you a clean, expandable PRD v0.1 that you can hand to designers, engineers, or AI collaborators without collapsing the magic.
Below is a starter PRD—intentionally scoped to first principles and MVP clarity, while leaving room for depth, ethics, and iteration.
Garden your heart.
Working Name: cori Alt Name: cœurculture Platform: Mobile (iOS / Android) + Web Status: Concept → Discovery → MVP Definition
cori is a reflective mental-health cultivation app that transforms spoken human experience into a living, evolving garden.
By recording and transcribing reflective conversations—solo or with another person—cori translates emotional tone, themes, and patterns into symbolic growth within a personal garden. Over time, users see not just what they say, but how they say it, and how their inner landscape changes.
Unlike productivity or symptom-tracking tools, cori emphasizes meaning, metaphor, and agency, allowing users to both witness their patterns and consciously reshape their inner terrain.
Think: FarmVille × voice journaling × emotional literacy But slow, reflective, and deeply human.
People struggle to:
- Recognize emotional patterns in their own speech
- Reflect on change over time without judgment or clinical framing
- Integrate intentional change alongside observed reality
- Stay engaged with mental-health tools that feel sterile, diagnostic, or punitive
Existing solutions focus on:
- Text-based journaling (high friction)
- Metrics-heavy mood tracking
- Therapist-only insights locked behind sessions
cori offers a living mirror, not a diagnosis.
- Reflective individuals who journal, meditate, or self-inquire
- People in therapy, coaching, or peer-listening relationships
- Creatives, facilitators, and emotionally literate users
- Coaches and therapists (as guides, not owners)
- Couples or trusted dyads practicing deep listening
-
User starts a session
-
Session can be:
- Solo (voice journal)
- Paired (coach, therapist, friend)
-
Audio is recorded (opt-in clarity is critical)
-
Session is transcribed
-
System analyzes:
- Topics (e.g., mother, work, body, joy)
- Emotional tone (e.g., deflated, energized, tense, warm)
- Frequency and recurrence
-
Concepts map to garden elements:
- Repeated themes → trees, patches, beds
- Tone influences health, size, light, or decay
-
Example:
- “Mom” spoken repeatedly with heaviness → withering tree
- “Children” spoken with warmth → growing sunflowers
Two modes:
-
Live Mode (optional)
- Garden evolves visually during the session
-
Post-Session Mode
- Garden is generated after transcription
- User can review and correct system misinterpretations
Users may:
- Tend the garden (gentle, non-gamified interactions)
- View history layers
- Add reflections or intentions
There is always a clear distinction between:
- Observed Reality (what speech + tone produced)
- Intentional Change (what the user manually adjusts)
This creates:
- Accountability without shame
- Desire without denial
- Truth without rigidity
Every element has:
-
Transcription Layer
- Immutable history
- What actually occurred
-
Intent Layer
- Manual edits
- Symbolic aspirations
- Conscious re-patterning
Visually, this could appear as:
- Ghost outlines
- Roots vs foliage
- Soil memory beneath surface growth
- Audio recording
- Transcription
- Theme + tone detection (basic)
- Garden generation
- Post-session review & correction
- Manual garden manipulation
- History vs intent distinction
- Longitudinal trend views
- Seasonal cycles
- Guided prompts
- Therapist / coach notes
- Exportable reflections
cori is not:
- A diagnostic tool
- A replacement for therapy
- A streak-based productivity app
- A “fix yourself” machine
The product must avoid:
- Clinical scoring
- Gamified shame
- “Optimization” framing of human emotion
- Clear consent before recording
- User owns their data
- Transparent explanations of how interpretations are made
- Easy correction of system misunderstandings
- No third-party data resale
- Explicit positioning as reflective, not authoritative
These should guide early research and prototyping:
- What metaphors resonate most across cultures?
- How literal vs abstract should visual feedback be?
- How much “accuracy” do users expect from tone analysis?
- What level of AI interpretation feels supportive vs invasive?
- How do users want to revisit painful growth safely?
- Users return without prompts
- Users correct the system instead of abandoning it
- Users describe insights in metaphor, not metrics
- Sessions feel held, not extracted
-
Write 1–2 canonical user stories
-
Sketch a single garden screen
-
Prototype:
- Audio → transcript → single plant
-
Define a tone taxonomy (very small at first)
-
Decide: live rendering vs post-session only (MVP choice)
If you want, next we can:
- Turn this into a Notion-ready PRD
- Define the plant / emotion taxonomy
- Draft a designer handoff brief
- Or write the “ethics & consent manifesto” that ships with v1
This is a rare idea, Brett. It wants to be grown carefully—just like the thing it’s about 🌱
- big inspiration from the style and feel of Terra Nil by FreeLives and Devolver https://www.terranil.com/