Current Project · Local-first AI reading product

eisonAI

eisonAI is not centered on chat. It places AI between capture, indexing, context, and retrieval, so saved material remains understandable after it leaves the original page.

AI ProductizationLocal-firstKnowledge ToolsSwift

Narrative

Design and engineering judgement

Product frame

The product works on reading friction: long-document routing, saving, context recovery, and privacy-aware AI use without turning the experience into a chatbot.

Engineering frame

The system combines app surfaces, Safari Extension, Cognitive Index, local data management, and CloudKit sync into one reading workflow.

Design frame

The interface story is structure before content: make the shape of information visible first, then make reading and retrieval easier.

System

Components and visibility

Component	Type	Visibility	Role
eisonAI app	AI reading product	Public product	Core experience for reading, saving, indexing, and retrieval.
Safari Extension	Capture surface	Product feature	Connects web reading to capture and indexing.
Cognitive Index	AI indexing concept	Public narrative	Generates structure and usage context around content.
Swift codebase	GitHub repo	Public	Readable entry point for Apple platform implementation.

Public URLs

Public surfaces

Product App Store

Public product surface for the local-first AI reading app.

apps.apple.com/us/app/eison-ai/id6484502399 Source GitHub

Readable implementation surface for Apple platforms, local AI workflow, and app structure.

github.com/qoli/eisonAI Media YouTube demo

Public video surface for seeing the AI reading workflow and product behavior.