OktoNote

The problem

Thoughts arrive in every format — a voice memo in the car, a photo of a whiteboard, a forwarded document, a quick line of text — and they scatter across a dozen apps. Capturing is easy; finding and organizing later is where it falls apart.

What OktoNote does

OktoNote is built around one idea: capture anything, and let AI do the organizing.

• Multi-modal capture. Text, voice, photos, and documents all go into the same place, with voice-to-text transcription built in.
• Ten knowledge card categories. The AI automatically sorts each capture into the right type — tasks, todos, journals, itineraries, links, attachments, lists, discussion notes, details to remember, and study materials.
• Feed and timeline views. A feed of organized knowledge cards plus a chronological timeline of your original source content.
• Intelligent search. Semantic search goes beyond keywords to find what you meant, not just what you typed.
• Works offline. Full offline functionality with cross-device sync.

How it’s built

OktoNote pairs a React Native app with an AI processing pipeline built on LangChain, using ChromaDB for vector embeddings and semantic retrieval. The backend handles authentication, sync, and multi-format note processing, with content categorization and extraction running through the AI layer.