Stella

Product Introduction

1. Definition: Stella is a local file search utility and AI-powered productivity tool for macOS. It functions as a semantic search engine that indexes and retrieves files based on the meaning of their content, not just filenames or metadata.
2. Core Value Proposition: Stella exists to solve the file discovery problem on local machines. It replaces frustrating, filename-dependent searches with intuitive, natural language search, allowing users to find documents by describing what they are about (e.g., "meeting specs," "exam cheat sheet"). Its core promise is to make file retrieval instant, private, and effortless.

Main Features

1. Feature Name: Semantic Document Understanding Stella employs an on-device AI model (specifically optimized for Apple Silicon) that reads and indexes the content of text-based files. Using natural language processing (NLP) and likely transformer-based architectures, it constructs a semantic understanding of each document. The technology works by analyzing text within PDFs, notes, spreadsheets, and documents to build a context-aware index. This allows it to map the conceptual meaning of a query to the relevant content, ignoring irrelevant filenames like finance_2024_v3.xlsx when searching for a "Q3 revenue report."

2. Feature Name: Privacy-First Architecture Stella is built on a completely local and private architecture. All processing, indexing, and search operations occur directly on the user's machine. There is no cloud component, telemetry, or account requirement. The system is designed with zero data leakage; files never leave the device, not even for processing. This uses a privacy-by-design approach, making it a secure solution for handling sensitive corporate documents, personal notes, or confidential academic materials.

3. Feature Name: Universal File Launcher Beyond search, Stella integrates a unified shortcut bar that serves as a powerful launcher. Once a file is found via natural language search, users can drag it directly into applications like Gmail, open it with its default app, or launch any other installed application from the same interface. This creates a streamlined workflow, reducing the cognitive load and time spent switching between Finder, Spotlight, and various applications.

Problems Solved

1. Pain Point: The "Spotlight Failure" - Traditional OS search tools like macOS Spotlight are inadequate for finding files when the user cannot recall the exact filename or when filenames are non-descriptive (e.g., Document1.pdf, final_v2_FINAL.docx). Stella solves the core problem of file retrieval based on content knowledge, not memory of naming conventions.
2. Target Audience: This product is essential for knowledge workers, students, researchers, writers, developers, and creative professionals who manage large volumes of local documents. It is specifically valuable for users with privacy concerns (e.g., lawyers, healthcare professionals, journalists) and power users seeking to maximize workflow efficiency on macOS.
3. Use Cases: Key scenarios include locating a research paper by recalling its thesis, finding a financial model by its purpose ("the budget forecast spreadsheet"), retrieving meeting minutes from a specific project, or surfacing creative briefs and design specs without browsing through hundreds of identically named folders. It is indispensable for digital clutter management and for professionals working off-grid or with strict data sovereignty requirements.

Unique Advantages

1. Differentiation: Compared to macOS Spotlight, which relies heavily on filenames and metadata, Stella performs deep content analysis. Unlike cloud-based search services (e.g., Google Drive search) or other desktop search tools, Stella offers superior privacy with its fully local, offline-capable operation. It requires no subscription, account, or internet connection, setting it apart from freemium or SaaS models.
2. Key Innovation: The primary innovation is the successful implementation of a capable, on-device AI model for consumer file search. By running a semantic indexing engine locally on Apple Silicon, Stella delivers cloud-level understanding without the privacy trade-offs or latency. Its ability to generate meaning from raw text locally and provide instantaneous results is a significant technical and user-experience advancement.

Frequently Asked Questions (FAQ)

1. What file types does Stella search? Stella indexes and understands text-based files, including PDFs, Word documents, text notes, spreadsheets (XLSX, CSV), and other common document formats. It extracts and analyzes the textual content to build its semantic search index.
2. How is Stella more private than Spotlight or cloud search? Stella processes all data 100% on your Mac. No file content or search queries are ever sent to a server, stored in the cloud, or used for telemetry. This makes it a fully private search engine, ideal for sensitive or confidential information.
3. Can Stella work completely offline? Yes, once installed and folders are selected for indexing, Stella works entirely offline. It does not require an internet connection for searching, making it reliable for use on airplanes, in remote locations, or in secure environments.
4. What are the system requirements for Stella? The current beta version requires macOS 13 Ventura or later and is optimized for and requires Apple Silicon (M1 chip or later). The download size is approximately 1.5 GB, as it includes the necessary on-device AI models for processing.
5. How does Stella's search differ from searching by filename or tags? Stella's search is intent-based and semantic. You search for the meaning or topic of a document ("project proposal"), not its name (Proposal_v3_final.docx). It understands synonyms and context, so you can find a document about "Q3 revenue" even if it is named finance_2024.xlsx.