char

This repository introduces Char, an AI notepad designed for meetings, aiming to streamline note-taking and summarization.

For the Non-Technical Reader

Imagine you're in a meeting. Instead of frantically typing every word, you jot down only the key points. Char is like a super-smart assistant that listens to the entire meeting, transcribes it, and then crafts a perfect summary based on your notes. It's like having a personal note-taker that works offline, ensuring privacy and efficiency. Think of it as a digital version of taking shorthand notes, but with AI doing the heavy lifting of creating a coherent summary afterward. This changes how users engage in meetings, allowing them to focus on the discussion rather than the documentation.

For the Technical Reader

Char is an AI notetaking application that transcribes meetings by listening to audio input from the user's computer. It is designed to create summaries based on user memos, with the option to run completely offline using LM Studio or Ollama for local LLM processing. The application stores notes, transcripts, and data locally in a SQLite database. It supports custom language models via Ollama or third-party APIs like Gemini, Claude, or Azure-hosted GPT. The application also offers note templates and AI chat functionalities. Integrations include Apple Calendar, Contacts, and Obsidian, with future integrations planned for Notion, Slack, Hubspot, and Salesforce. The macOS version is currently in public beta, with Windows and Linux versions planned for Q2 2026.

Why It Matters

Char's focus on local processing and offline capabilities addresses privacy concerns, making it suitable for environments where data security is paramount. The ability to bring your own LLM allows organizations to maintain compliance with their specific requirements. By providing both cloud-powered and offline options, Char balances convenience with data control. This lowers the barrier to entry for AI-driven note-taking, especially in privacy-sensitive industries.

The "Voice AI Space Lab" Idea

Imagine building a "Meeting Insights Dashboard" using Char. You could combine Char's meeting summaries with sentiment analysis to gauge team morale during different project phases. By integrating this with project management tools, you could automatically identify potential roadblocks based on the tone and content of meeting discussions, offering proactive alerts to project managers. This goes beyond simple note-taking and enters the realm of predictive project management.