Creative OS is an AI-powered desktop application designed for writers, filmmakers, musicians, and content creators. Unlike a traditional chatbot, it serves as an intelligent operating system for creative work by indexing local creative assets—including Final Draft screenplays, PDFs, lyrics, notes, outlines, shot lists, and research documents—and enabling Claude to search, reason, analyze, and edit them through Retrieval-Augmented Generation (RAG).
The platform is fully local-first, ensuring user privacy while providing AI-assisted writing, semantic search, knowledge retrieval, document editing, and long-term contextual memory. Built using Electron and React, Creative OS offers a native desktop experience with a modular architecture that can be extended through MCP (Model Context Protocol) servers.
Project walkthrough
Demo Video
What we built
Local Knowledge Indexing
Built document ingestion pipelines to index local creative assets including Final Draft (.FDX), PDFs, lyrics, notes, outlines, and research documents.
Implemented automatic metadata extraction and searchable document indexing.
Enabled semantic retrieval across an entire local creative knowledge base.
Retrieval-Augmented Generation (RAG)
Designed and implemented a complete RAG pipeline for contextual document retrieval.
Generated vector embeddings and stored them in a local vector database.
Retrieved relevant document context before every AI request, improving response accuracy and grounding.
Claude AI Integration
Integrated Claude API to provide AI-powered writing assistance.
Combined Claude responses with retrieved document context for highly personalized outputs.
Final Draft (.FDX) Editing
Impact
Centralized an entire creative knowledge base into a single AI-powered desktop workspace.
Eliminated manual searching across hundreds of creative documents through semantic retrieval.
Enabled Claude to reason across multiple files while maintaining contextual accuracy using RAG.
Accelerated screenplay writing, research, brainstorming, and content creation through intelligent AI-assisted workflows.
Delivered a scalable local-first architecture that combines document management, semantic search, AI editing, and future MCP extensibility into one platform.
