The questions that actually decide it.

That's exactly who it's for. Sussur is built for the one to five people selling at a software company with no sales manager, no call-review tool, and no enablement team — the first sales hire, and the founder still closing their own deals. Nothing about it needs a team to be worth having.

A Mac with Apple Silicon (M1 or later) running macOS 13 or newer. Speech-to-text runs on the machine itself, which is why the requirement is firm rather than a suggestion. There's no web app and no Windows build today.

Your audio and video never leave your device — capture and speech-to-text run on your Mac, and no recording is uploaded. To generate the coaching, transcript text goes to the model provider you configured, on your own API key; that is the one thing that leaves, and we'd rather say it plainly than let a privacy headline imply otherwise. Point Sussur at Ollama on the same Mac and even that stays local.

No. Sussur does not infer emotion from faces, expressions, or vocal tone — not yours and not the other person's. The coaching is drawn from the transcript, as text, and nothing else.

Sussur is a panel on your screen — no bot joins the meeting and nothing changes about how you take the call. We do not build a mode whose purpose is to hide it from the person across the table, and we won't: coaching you is the product, concealing it is not. You're responsible for the recording and consent laws where you operate, and telling people is usually the easier conversation anyway.

Sussur captures system audio on the Mac, so it works with Zoom, Google Meet, Microsoft Teams, phone calls, and any other audio source — no per-platform integration to set up.

Yes — Sussur runs the coaching on your own key. Paste an OpenRouter API key and pick a model, or point it at Ollama running on the same Mac if you'd rather nothing left the machine at all. You pay your provider directly for what you use; we never see that key or that bill.

Nothing. It is free while it is in development. There is no account to create and no charge of any kind from us. The only cost is whatever your own AI provider bills you for the coaching requests — or nothing at all, if you run a local model with Ollama.

Not yet. Sussur is Mac-first because the speech-to-text that runs on-device relies on Apple Silicon. The Windows port is a 6-8 week effort sized for the priced seed at month 13-18.

The product is tuned for high-stakes B2B sales calls, where a single conversation moves a deal. People do use it for adjacent conversations — investor pitches, key client check-ins, vendor negotiations — but that's where it's sharpest.