Getting started

RepCue is a desktop copilot plus a web console. Reps run the desktop app during calls; admins and managers use the console for knowledge, analytics, coaching and settings. This page gets a team from zero to the first live suggestion.

1. Create the workspace

1. Go to repcue.io/signup and pick Founding or Team. Both start with a minimum of three seats.
2. Enter the organisation name, your name and email, and the seat count. A card is required before access; there is no free tier.
3. Checkout opens in a hosted page. When it completes you land in the console, signed in as the first admin.

Invite the rest of the team from Settings → People. Invites are emailed; each one consumes a seat when accepted.

2. Install the desktop app

Download from repcue.io/download. Builds:

The app updates itself from the same host the download links point at. IT can pin a release with the MDM notes on the download page.

3. Grant permissions

The desktop app captures two channels: your microphone (you) and system audio (everyone else). Separating them at the source is why RepCue never has to guess who is speaking.

macOS

• Microphone: the OS prompts on first run. Granting it shows a live meter in the wizard so you can see exactly what the permission enables.
• System audio: RepCue uses the documented screen-and-audio capture API, so macOS lists it under Privacy & Security → Screen & System Audio Recording. The wizard deep-links to that pane. A freshly granted permission only takes effect after the app restarts; the wizard tells you when that's needed.

Windows

• Microphone: granted through the standard Windows microphone privacy setting.
• System audio: Windows exposes loopback capture of the default output device to any application; no elevated consent, admin rights or driver is required. Security teams often ask about this. It's the same mechanism screen recorders and meeting apps use.

Neither platform needs a virtual audio driver, and nothing is installed at the kernel level.

4. Acknowledge the consent policy

Before the first call, each rep confirms they understand their disclosure obligations. Your admin sets the per-jurisdiction policy (record, disclose, confirm, or assist-only) in Settings → Consent; the app follows it automatically. See Consent & recording policies.

5. Run the roleplay

The wizard ends with a roleplay against an AI customer. With no documents loaded it demonstrates the HUD; once your knowledge base is indexed, re-run it from the HUD (Roleplay) and the customer argues from your pricing and your competitors. See Roleplay mode.

6. Load knowledge

Drop your pricing sheet, playbook, battlecards, case studies and FAQ into Console → Knowledge, or connect Notion, Google Drive or Confluence. Parsing, chunking and embedding take about two minutes per document. Every suggestion from then on cites the document and section it came from.

7. Connect the stack (optional)

Settings → Integrations connects Salesforce, HubSpot or Pipedrive (deal context in, approved changes out), Google or Microsoft (calendar auto-arm, pre-call brief, send follow-ups as the rep), Slack (summaries to deal channels) and Recall (meeting bot for machines without the app). See Integrations.

8. First call

Open the HUD before the meeting, or let the calendar arm it two minutes before. The status rail reads Listening · both channels when both legs are live. Cards appear on the right edge of your meeting window; ⌘⇧1/2/3 expands them, ⌘⇧H hides everything instantly.

When the call ends the HUD flips to the review sheet: summary, CRM changes, follow-up email, notes. Edit anything, then Ship, or open the call in the console.