Five concrete differences
Cue
Granola
Where the audio goes
Stays on your Mac. Transcription runs on-device through macOS SFSpeechRecognizer. No audio file leaves the machine.
Audio is sent to a third-party cloud for transcription, governed by Granola's vendor terms.
Speaker separation
You and Others, on separate streams. Mic and system audio are captured and labeled independently — so when a client commits to a date, the transcript shows who said it.
A single merged transcript without per-speaker attribution.
AI infrastructure
FuelIX — TELUS Digital's sanctioned AI platform. Same governance and data boundary as the rest of WillowTree's enterprise AI stack.
Third-party AI providers chosen by the vendor, outside WillowTree's enterprise governance.
Answers during the meeting
Proactive answers, mid-sentence. Cue listens for questions, searches your knowledge base — proposals, past meetings, GitHub repos — and surfaces an answer before the client finishes asking.
Notes and summaries land after the meeting ends.
Where meetings live
On your machine. Transcripts, summaries, and chat history are stored in a local SQLite database and exported as Markdown into a folder you choose.
Meetings are stored in Granola's cloud, accessed through their app.

Comparison reflects publicly documented Granola behaviour as of April 2026.

Built for the rooms you actually work in.

If you run client engagements on a TELUS Digital MacBook, Cue is for you.

Download for Mac