Booth Conversation Capture

Download the LeadFuel desktop app from the Download page (lead-fuel.com/download).

1. 1Pick your build: Apple Silicon, Intel, or the Windows 64-bit installer
2. 2On a Mac, open the DMG and drag LeadFuel to Applications. On Windows, run the installer
3. 3Launch it. macOS may ask you to approve it in System Settings → Privacy & Security, and Windows may show a SmartScreen notice where you click More info then Run anyway
4. 4Grant microphone access when prompted

The app will download the AI transcription model on first launch (~500MB). This only happens once and may take a few minutes depending on your connection.

Open the app and go to the Settings tab:

1. 1API URL: Enter https://lead-fuel.com
2. 2API Key: Enter your API key from Dashboard → Settings → Integrations → API Keys
3. 3Click Test Connection to verify the connection works

If the test succeeds, you'll see a green checkmark. If it fails, double-check the API URL (no trailing slash) and make sure your API key is correct and active.

The API key authenticates the desktop app with your LeadFuel account. Each key is tied to a specific tenant and user.

The microphone is the easy part. What decides whether a transcript becomes a good follow-up is what your reps actually say.

Say the visitor's name and company out loud, early in the conversation. LeadFuel matches a conversation to a person by listening for the name. If nobody says it, the analysis has a transcript and no idea whose it is — at a known-attendee event it may not link to anyone, and at a busy badge event LeadFuel will hold the draft for review rather than guess between two visitors.

It sounds like ordinary manners, which is the point: - "Hey George, great to meet you. How long have you been at Moen Faucets?" - "Mr. Smith, good to meet you. Tell me about how long you've been at Moen." - "So Priya, what brought you over to the booth?"

Scan when you meet them, then talk. LeadFuel takes the 5 minutes before the scan and the 10 after it, so scanning at hello puts the whole conversation inside the window. It is also what brings up the Instant Insights sheet the rep reads while they are still talking.

Add a note before moving on. Rep notes outrank everything else in the follow-up draft. Five words will do.

Plug in your microphone (USB or wireless receiver) and go to the Settings tab in the app.

The app will show a list of available audio input devices. Select your microphone from the dropdown. If you're using a wireless system like the Hollyland Lark M2S or Lark Max 2, select the USB receiver — it will appear as a single audio device regardless of how many transmitters are connected.

For wireless clip-on mics, clip the transmitter at chest level about 6–8 inches from the speaker's mouth.

On the main screen, you'll see an event selector. Choose your event from the dropdown, or create a new one directly in the app.

The event name should match what you created in the web dashboard. If you type a name that doesn't exist, the app will create it — and it will appear in your web dashboard when the transcripts sync.

All conversation transcripts captured during a session will be associated with this event. When badge scans happen on the web side, the system matches them to conversations from the same event.

Click the Start button to begin capturing. The app will:

1. 1Create a new session (named with the current date and time)
2. 2Begin capturing audio from your selected microphone
3. 3Transcribe the audio in real-time using AI running locally on your Mac
4. 4Display the live transcript as text on screen

The transcription happens entirely on your device — no audio data is sent to the internet. The AI model uses your Mac's GPU for acceleration, so transcription is near real-time on Apple Silicon (M1/M2/M3/M4) Macs.

While the session is active, you'll see:

• Live audio levels — the dB meter shows real-time input volume
• Live transcript — text appears as the AI processes the audio
• Session duration — how long the current session has been active
• Connection status — whether the app can reach the cloud (not required for capture, only for sync)

Check the app periodically to make sure: - Audio levels are still good (microphone hasn't been disconnected or muted) - The transcript is updating (the AI is processing correctly) - The laptop is plugged in (GPU transcription uses battery)

You can continue scanning badges on your phone while the desktop app captures conversations — they're completely independent.

Transcripts sync to your LeadFuel cloud dashboard automatically when your laptop has an internet connection. This happens in real-time if you're connected, or when you reconnect after being offline.

The sync sends only text transcripts — no audio data ever leaves your device.

To verify sync status: 1. Check the connection indicator in the app (green = connected, yellow = syncing, red = offline) 2. Open your web dashboard → Events → [Your Event] → Capture Sessions 3. You should see the session listed with transcript chunks

If you were offline during the event, the transcripts will queue up and sync in bulk when you reconnect. This can take a few minutes for a full day of conversations.

At the end of the day (or when you're done capturing), click Stop to end the session.

The session is saved locally and will continue syncing if there are any remaining transcript chunks that haven't been uploaded.

You can start a new session the next day — each session is tracked separately with its own timestamps, making it easy to identify which conversations happened on which day.

After the event, you can safely close the app. All data is preserved locally and in the cloud (once synced).