Note67 - Private Meeting Notes Assistant

A second critical feature is speaker distinction, which automatically separates the user's voice from other meeting participants by leveraging both microphone input and system audio capture. This capability is essential for generating clear, organized transcripts where different speakers are identified, making the notes much more readable and actionable. The system uses intelligent audio processing to differentiate between the local user speaking and voices coming from video conferencing applications like Zoom or Teams. This allows Note67 to create a coherent transcript that attributes dialogue correctly, even in hybrid meeting environments where some participants are remote. By accurately distinguishing speakers, the tool saves users the tedious manual work of labeling who said what, streamlining the post-meeting review process and ensuring the transcript reflects the actual conversation flow.

The third feature group encompasses private AI summaries generated using Ollama and local large language models (LLMs). After transcription, Note67 can process the text to produce concise, intelligent summaries that highlight key points, decisions, and action items, all without any cloud dependencies. Users have full control over their AI stack, as they can choose and configure any local LLM compatible with Ollama, tailoring the summarization to their specific needs or preferences. This feature transforms raw transcripts into immediately useful insights, enabling users to quickly grasp meeting outcomes without rereading lengthy text. Since the LLM runs locally, the summary generation is as private as the transcription itself, ensuring that even the condensed notes never leave the device, maintaining end-to-end confidentiality throughout the entire note-taking workflow.

Overall, Note67 works by integrating several technical components into a seamless, native application built with the Tauri framework using Rust and React. It captures audio through system-level APIs like ScreenCaptureKit on macOS and equivalent Windows interfaces, then processes that audio locally using Whisper for transcription. The application employs Voice Activity Detection (VAD) and echo deduplication algorithms to filter out background noise and speaker echo, especially when using speakers instead of headphones, ensuring clean audio input. Transcripts are managed and stored in a local SQLite database, with the user interface built in React and TypeScript providing a clean, responsive experience. All AI operations, including summarization via Ollama, are executed on-device, leveraging the local computational resources without requiring an internet connection, which defines its core technical approach as offline-first and privacy-centric.

Users benefit from measurable outcomes such as enhanced productivity, as they can focus fully on the meeting discussion instead of frantic note-taking, leading to better engagement and collaboration. The automatic, accurate transcripts reduce the time spent on manual documentation, allowing teams to allocate more resources to strategic tasks rather than administrative overhead. From a security perspective, organizations achieve compliance with data protection regulations like GDPR or HIPAA by ensuring meeting content never leaves their controlled environment, mitigating legal and reputational risks. The local processing also means there are no subscription fees or ongoing costs after purchase, providing a clear return on investment through one-time licensing and eliminating unpredictable cloud service expenses over time.

Concrete use cases include legal professionals who need to transcribe client consultations without violating attorney-client privilege, as Note67 keeps all case discussions securely on their laptop. Healthcare providers can document patient interactions for accurate records while maintaining HIPAA compliance, since no protected health information is uploaded to external servers. Corporate strategy teams can capture brainstorming sessions about mergers or product launches, confident that their intellectual property remains internal. Remote workers participating in daily stand-ups or project reviews can generate instant summaries to share with colleagues, streamlining communication. Executives in board meetings can rely on the tool to document decisions and action items privately, ensuring sensitive corporate governance discussions are not exposed.

The target users are professionals across industries where meeting privacy is critical, including lawyers, doctors, financial advisors, consultants, and corporate executives. It also appeals to developers, IT administrators, and privacy advocates who prefer open-source, transparent tools. Note67 integrates with the user's existing meeting setup by capturing system audio from any application, such as Zoom, Microsoft Teams, Google Meet, or even in-person conversations via microphone. Its tech stack includes React, TypeScript, Tailwind CSS for the frontend; Rust, Tauri, and SQLite for the backend; and whisper-rs, Ollama, and local LLMs for AI capabilities. Pricing is based on a one-time purchase model with no subscriptions, available for Windows 10+ and macOS 13+ as a native desktop application, emphasizing ownership and long-term value.

In summary, Note67 delivers a fundamentally different approach to meeting notes by placing privacy at the forefront without sacrificing functionality. It empowers users to leverage cutting-edge AI for transcription and summarization while retaining complete data sovereignty, addressing the growing demand for tools that respect confidentiality in an era of pervasive cloud surveillance. By running entirely on-device, it provides peace of mind that sensitive discussions remain secure, making it an indispensable asset for any professional who values both efficiency and discretion in their communication workflows.

Updated 2026-02-28