Context capture
Words and phrases keep source, sentence and user intent.
Introduction
Inside Kotori
AI language tools easily become chat toys unless they preserve memory, retrieval practice and learner intent.
What had to stay educational
The UX is light during capture and focused during review: users should stay in the language, not in administration.
SRS
retention loop
AI
context assistant
Capture
from real material
Practice
active recall
Features
Product features with technical depth
We centered the product on the learning loop: capture, explain, schedule, recall and use.
AI explanations
Assistance explains usage without hiding uncertainty.
Spaced repetition
Review is connected to real material instead of isolated cards.
Progress evidence
The product shows retention and usage, not only streaks.
AI content pipeline
Capture, explanation and review can use model output while preserving learner control.
Capture sheet
A quick capture flow turns a phrase into a reviewable memory.
Review room
Cards include context, hints and recall feedback.
Immersion dashboard
Users see what material is feeding their learning loop.
Challenges
Disconnected practice
Vocabulary lists, notes, immersion and speaking practice usually require separate tools.
Technical challenges
AI uncertainty
Explanations had to help without hiding ambiguity or replacing active learning.
Retention loop
Captured language needed to become durable review, not another forgotten note.
Capture friction
The flow had to stay light enough for real immersion contexts.
Domain challenges
Disconnected practice
Vocabulary lists, notes, immersion and speaking practice usually require separate tools.
Capture sheet
A quick capture flow turns a phrase into a reviewable memory.
Review room
Cards include context, hints and recall feedback.
Technology stack
Technology stack behind the product
Kotori combines content capture, AI processing, vocabulary memory, SRS scheduling and practice flows over a privacy-conscious learning backend.
Backend
.NET learning services
AI
LLM explanations, summaries
Data
Vocabulary memory, review state
Clients
Web and mobile learning surfaces
Web app
Angular, capture and review surfaces
Mobile
.NET MAUI direction, learning sessions
Cloud
GCP, Kubernetes, AI workers
.NET learning services
summaries
review state
Angular
.NET MAUI direction
GCP
AI workers
AI
Clients
Mobile
Language
Immersion
LLM explanations
Vocabulary memory
Web and mobile learning surfaces
capture and review surfaces
learning sessions
Kubernetes
Backend
Data
Web app
Cloud
SRS
LLM explanations
Vocabulary memory
Web and mobile learning surfaces
capture and review surfaces
learning sessions
Kubernetes
Backend
Data
Web app
Cloud
SRS
.NET learning services
summaries
review state
Angular
.NET MAUI direction
GCP
AI workers
AI
Clients
Mobile
Language
Immersion
Preview
Platform insight
Selected product screens showing information architecture, UX decisions and real system surfaces.
Japonics
Make AI language learning measurable without making it mechanical
Kotori is a craft study in educational UX and responsible AI assistance.