SuperBrain

Product Introduction

1. Definition: SuperBrain is an open-source, self-hosted "Personal Knowledge Management" (PKM) and AI-augmented content archive system designed specifically for the Android ecosystem. Technically categorized as a full-stack application, it comprises a React Native (Expo SDK 54) mobile frontend and a Python 3 (FastAPI) backend that serves as an intelligent intermediary between raw web URLs and structured, searchable data.

2. Core Value Proposition: SuperBrain exists to eliminate "content silos" and the "saved content graveyard" phenomenon. It provides a privacy-first, subscription-free alternative to cloud-based bookmarking apps by automating the extraction, summarization, and categorization of URLs from platforms like Instagram, YouTube, and various web domains. Its primary objective is to facilitate "smart discovery" and "data sovereignty," ensuring that users own their intelligence data in a local SQLite database without exposure to data harvesting.

Main Features

1. Multi-Provider AI Model Router: SuperBrain implements a sophisticated, self-optimizing AI routing engine. To ensure high availability and bypass rate limits of free-tier APIs, the system uses an Exponential Moving Average (EMA) ranking algorithm to track response times across Groq, Google Gemini, OpenRouter, and local Ollama instances. If a primary provider (like Groq) fails or hits a rate limit, the router transparently falls back to the next fastest available model, ensuring that content analysis never stalls.

2. Automated Multimedia Processing & Transcription: Unlike traditional bookmarking tools, SuperBrain performs deep analysis of media. It utilizes OpenCV for frame extraction in vision-based models, the Groq Whisper API (with local OpenAI Whisper fallback) for high-accuracy audio transcription, and the Shazamio API for background music identification within Instagram Reels. This allows users to search for content based on spoken words or specific background tracks.

3. Native YouTube and Social Media Intelligence: The backend features specialized scrapers and analyzers, including native YouTube understanding via the Gemini API (which analyzes video content via URL without requiring full downloads) and Instaloader/Instagrapi for Instagram content. For general web pages, it employs a multi-strategy scraping approach using newspaper4k, trafilatura, and the Wayback Machine to ensure content is captured even if the original source is modified or removed.

4. Offline-First Architecture and Smart Notifications: The mobile application is built with an offline-priority mindset, utilizing an internal queueing system that saves URLs locally and synchronizes with the self-hosted server once a connection is established. It features a advanced Android notification system that schedules "Watch Later" reminders and "Urgent Alerts" for deadline-sensitive content, helping users actually consume the information they save.

Problems Solved

1. Fragmented Information Retrieval: Users often save content across disparate platforms (Instagram Saves, YouTube Watch Later, Chrome Bookmarks) where it becomes unsearchable and forgotten. SuperBrain centralizes these into a single, full-text searchable repository.

2. Privacy Concerns and Data Monetization: Most "second brain" applications are SaaS-based, meaning user data is stored on third-party servers and often used for advertising profiles. SuperBrain solves this through self-hosting and local SQLite storage (WAL mode), giving the user 100% control over their data.

3. Target Audience:

• Privacy Advocates: Individuals who want to avoid Big Tech data harvesting.
• Students and Researchers: Users needing to archive and summarize academic or technical content from various web sources.
• Digital Content Creators: Marketers and creators who need to track trends, music, and transcripts across social media platforms.
• Tech Enthusiasts: Developers looking for a customizable, self-hosted AI stack.

4. Use Cases:

• Archiving a recipe from a fleeting Instagram Reel for later search.
• Summarizing a 2-hour YouTube lecture into a 2-minute readable brief.
• Building a private library of technical articles that remains accessible even if the original site goes offline.
• Identifying a song from a background of a social media post without leaving the app.

Unique Advantages

1. Zero-Cost Maintenance: By leveraging the generous free tiers of providers like Google Gemini and Groq, or running models locally via Ollama, SuperBrain allows users to maintain a high-performance AI system with zero monthly subscription fees.

2. Hybrid AI Strategy: It bridges the gap between cloud speed and local privacy. Users can use cloud APIs for fast processing when online and switch to a fully local, air-gapped setup (Ollama + Local Whisper) for maximum privacy.

3. Advanced Developer Experience: The project includes a dedicated CLI (superbrain-server) that automates backend deployment via npm, making it accessible to users who are not experts in Python environment management or Docker.

Frequently Asked Questions (FAQ)

1. Is SuperBrain truly private? Yes. SuperBrain is a self-hosted solution, meaning the database (SQLite) resides on your own hardware (PC, Raspberry Pi, or private VPS). No data is harvested or sent to a central SuperBrain server. AI processing occurs via your chosen API keys or locally via Ollama.

2. Which AI models are required to run SuperBrain? SuperBrain is designed for flexibility. While it supports Groq, Gemini, and OpenRouter, you can start with just a single free API key from Google Gemini. For users seeking 100% offline functionality, it supports Ollama for text/vision and local OpenAI Whisper for transcription.

3. Can I save content from private Instagram accounts? Yes, provided you follow the account. SuperBrain includes an Instagram login utility that allows you to use your own credentials (stored locally) to authenticate requests via Instaloader, enabling the archival of posts from private profiles you have access to.

4. How do I access my SuperBrain from outside my home network? The system supports multiple hosting options. You can use an Ngrok tunnel (integrated into the superbrain-server CLI) for easy remote access, or deploy the backend via Docker on a VPS for a permanent, public-facing personal API.