note.md

Product Introduction

1. Definition: note.md is a local-first, privacy-focused knowledge management and academic writing application for macOS. Technically, it is a markdown-native desktop application that functions as an integrated research and writing environment, combining a reading studio, a knowledge graph, and local AI-powered semantic search.
2. Core Value Proposition: It exists to provide academics, researchers, and serious writers with a private, distraction-free workspace for deep focus and structured thought. Its primary value is enabling seamless synthesis of information from PDF sources into well-structured, interconnected writing, all while ensuring user data never leaves their local machine.

Main Features

1. Reading Studio: This feature provides a dual-pane interface for side-by-side reading and note-taking. Users can view a PDF in one pane and a linked markdown note in another. It enables smart citations, allowing users to highlight text in a PDF and automatically generate a properly formatted BibTeX or APA citation linked to that highlight within their note. The workspace supports GitHub Flavored Markdown, complex tables, and LaTeX math rendering.
2. Local Knowledge Graph (Graph View): The application automatically builds a visual, bidirectional link graph of all notes and imported sources (PDFs). This graph view displays relationships between concepts, articles, and reference materials, facilitating the discovery of non-obvious connections. It supports exporting the graph structure in a Neo4j-friendly format for advanced external analysis.
3. Local AI & Semantic Search (Premium Feature): Powered by on-device AI models, this feature allows users to search their entire workspace by meaning, not just keywords. The underlying architecture involves local embedding models to create vector indexes of text chunks, coupled with hybrid ranking algorithms that fuse semantic similarity with traditional keyword matching. This enables retrieval of relevant passages even when the exact query wording isn't present in the source material.
4. Source Management & Bibliography Export: Users can import scientific papers and complex PDFs directly into the application. The tool allows for manual or metadata-assisted filling of source information (author, title, etc.). A key function is the ability to export all managed sources as a single, consolidated .bib (BibTeX) file for use with LaTeX or other reference managers, streamlining academic workflow.
5. Matrix Extraction & Evidence Scan (Premium AI Features): These are advanced local AI pipelines. Matrix extraction automatically identifies and extracts structured data like figures, charts, and tables from PDFs for insertion into notes. Evidence Scan allows a user to input a claim or hypothesis; the system then scans the locally indexed corpus to find supporting or contradictory evidence across all imported sources and notes.

Problems Solved

1. Pain Point: Fragmented research workflows. Academics often juggle separate tools for PDF management (Zotero, Mendeley), note-taking (Obsidian, Notion), and writing (Word, Scrivener), leading to friction and lost context.
2. Target Audience: Primary personas include academic researchers, PhD students, nonfiction authors, and technical writers who work extensively with PDFs and require rigorous source citation and knowledge synthesis.
3. Use Cases: Writing a literature review by connecting insights across dozens of papers; developing a book or long-form article where maintaining source integrity is critical; building a personal knowledge base (a "second brain") from research materials without relying on cloud services.

Unique Advantages

1. Differentiation: Unlike cloud-based note-taking apps (Notion, Evernote) or reference managers, note.md is strictly local-first and markdown-native. Unlike other markdown editors (Obsidian, Typora), it offers deeply integrated PDF reading, citation management, and local AI features out-of-the-box, creating a unified, purpose-built environment for research writing.
2. Key Innovation: Its commitment to fully local AI processing. While some competitors offer AI features, they typically rely on API calls to cloud services (OpenAI, Anthropic). note.md performs embedding, semantic search, and data extraction entirely on the user's Mac, ensuring absolute privacy, eliminating data transfer costs, and allowing offline functionality. The platform's transparency in documenting its AI pipeline's chunking, ranking math, and model parameters is also a significant trust and innovation factor.

Frequently Asked Questions (FAQ)

1. Is note.md free to use? Yes, the core application for individual creators is completely free, including the Reading Studio, Markdown editing, Graph View, and basic source management. The premium local AI features (Semantic Search, Matrix Extraction, Evidence Scan) require a separate license.
2. Where is my data stored when using note.md? All data, including your notes, imported PDFs, and the AI-generated vector indexes, is stored exclusively in a local "vault" on your macOS machine. No data is sent to external servers, ensuring complete privacy and ownership.
3. How does the local semantic search work without the internet? The search uses on-device machine learning models (specifically, local embedding models) to convert your text into numerical vectors. These vectors are stored and compared locally on your Mac using efficient similarity search algorithms, requiring no cloud processing or internet connection.
4. Can I use note.md for collaborative writing projects? note.md is primarily designed as a private, local-first tool for individual deep work. It does not have built-in real-time collaboration features. Collaboration would involve sharing exported markdown files and .bib files, which aligns with its philosophy of using open, portable formats.
5. What markdown and citation formats does note.md support? The editor supports GitHub Flavored Markdown (GFM), including task lists, tables, and strikethrough. For citations and academic writing, it directly supports LaTeX math equations and can import/export bibliographies in the BibTeX format, with automatic generation of APA and BibTeX citation styles from PDF metadata.