AI LabsNEXUS AI Labs
Coming soon
Bob solution cover image
Step 01 — Meetings · Managing

Bob

Perfect meeting notes, with not a single character leaving your device

Back to solutions

100% on-device AI meeting copilot — transcription, speaker ID, summaries

macOS desktop (Wails v2)Wails v2100% local processingSpeaker diarization

When a meeting starts, Bob transcribes in real time, tells speakers apart, and when it ends, the minutes and a one-line summary are already done. FunASR SenseVoice speech recognition, CAM++ speaker embeddings, Qwen3 4B summarization on llama.cpp — all of this AI runs inside your Mac with no cloud. The only kind of meeting AI you can use without worrying about security.

Bob
Bob main screenshot
Bob screenshotBob screenshot

Key features

  • Real-time transcription + automatic speaker separation, re-identification via saved profiles
  • FunASR SenseVoice speech recognition · CAM++ speaker embeddings
  • Local Qwen3 4B summarization and suggestion synthesis on llama.cpp
  • CoreAudio microphone capture with a built-in Python speech server
  • Automatic setup on first launch (~1.5GB model download)

Why It Stands Out

Zero Cloud

Speech, transcription, and summaries all stay on your Mac. Meeting content never leaves your device — no conflict with corporate security policies.

Real-time Speaker ID

CAM++ embeddings automatically identify speakers and save name tags. The same voice is instantly recognized in your next meeting.

Complete Minutes, Instantly

The moment a meeting ends, a local Qwen3 4B model writes the summary, action items, and a one-line conclusion — ready to share without edits.

One-time Setup, Done

The Python environment and AI models install automatically on first launch. No additional configuration needed after that.