LifeOS

Problem

Knowledge captured from articles, videos, and notes tends to become unsearchable and disconnected the moment it's saved.

What I built

An AI-powered second-brain platform that ingests multimedia content, extracts and organizes key insights, and enables semantic search, knowledge discovery, and personalized recall.

System & workflow

An ingestion pipeline processes incoming content and extracts key insights, which are indexed in Supabase and made searchable through OpenAI embeddings, surfaced through a Next.js/TypeScript frontend.

My contribution

Built the full stack — ingestion, extraction, storage schema, and the semantic search interface.

Challenges & decisions

Structuring extracted insights so they stay useful and searchable instead of collapsing into another unsorted content dump.

Results

Live at life-os-xi-azure.vercel.app.