Inyeon
A journaling practice built around one rule: your inner world belongs to you.
Inyeon is a compassion journaling practice built on OFNR: Observations, Feelings, Needs, Requests. Your data never leaves your device. Not a policy. A structural guarantee.
People come to this page from different directions. Tell me who you are and I'll show you what's most relevant.
Recruiter View
You're evaluating product instinct and values-driven design. Inyeon shows both: a product where the success metric runs counter to every engagement playbook.
- Problem selection: chose compassion journaling specifically because the market optimizes for engagement. Inyeon optimizes for graduation: users who need it less.
- Framework application: NVC's OFNR (Observations, Feelings, Needs, Requests) turned into a data model. Psychology as schema design.
- Architecture as values: smart-client, dumb-server. Privacy isn't a policy, it's structural. Every journal entry lives in local SQLite. Server is stateless.
- Co-founder dynamic: built with Keiko, whose practice shaped the compassion arc from self inward to the more-than-human world.
- Degrowth metrics: defined and shipped a success measure where engagement going to zero means the product worked.
Investor View
You're reading to understand founder thinking and market positioning. Inyeon sits at the intersection of the mental wellness market and the sovereign AI thesis.
- Market timing: mental wellness app market is overcrowded with engagement-maximizing tools. The degrowth angle is a genuine category differentiation.
- Sovereignty as moat: on-device inference means no data portability concerns, no HIPAA exposure, no vendor lock-in for the user. Structurally aligned with the regulatory direction of travel.
- The graduation metric: a product that measures its own obsolescence is unusual. The business model question is real and worth a conversation.
- Firecracker isolation: per-user microVM at ~14 cents per user per month. Infrastructure cost is known and bounded at early scale.
- Co-founder: Keiko brings the contemplative practice depth. I bring the product and systems architecture.
Engineer View
You're here for the architecture. The interesting constraint is that the privacy guarantee has to be structural, not policy. That shapes every technical decision.
- Smart-client, dumb-server: journal entries, OFNR extractions, and compassion patterns all live in local SQLite. Server is stateless. No user data crosses the wire.
- OFNR extraction: on-device NLP extracts Observations, Feelings, Needs, and Requests from free-text journal entries. Model runs locally.
- Firecracker isolation: each user gets their own microVM via Firecracker. Sovereign compute at ~14 cents per user per month. No cross-user data surface.
- CQL v3.0: Compassion Query Language, a structured schema for representing compassion practice patterns over time. Journaling choices as the data model.
- Degrowth metrics pipeline: the system tracks its own usage and surfaces patterns that indicate the user is internalizing the practice. Success = lower engagement.
Just here as a person
No pitch. The teaser above is the whole thing in one sentence. Here's the longer version.
- I built this because I wanted a journaling practice that didn't surveil me. Most apps need your data to function. Inyeon doesn't.
- The OFNR framework comes from Nonviolent Communication. It turns "I'm frustrated" into something more precise: what I observed, what I felt, what I needed, what I'd ask for.
- The graduation metric is personal. I want tools that make me less dependent on them. That's the design goal.
- Keiko is the reason the compassion arc is real and not just a product frame. The outward expansion from self to others to the world came from her practice.
Different angle
The views above cover the most common approaches. If you're here for a different reason, direct conversation is faster.
- Tell me what brought you here and what you need to know.
- I'll respond with what's actually relevant.