Transcriber Configuration

Learn how to configure speech-to-text transcribers in Convocore Voice Suite. Transcribers convert user voice inputs into real-time text using providers like Deepgram.

---

​

🛠 Configuring the Transcriber

Follow these steps to set up the transcriber:

1. Navigate to the Voice Setup page in your Convocore dashboard.
2. Select the Transcriber tab.

---

​

🔧 Step 1: Choose Your Transcriber Provider

1. Use the dropdown to select your preferred provider.

---

​

🤖 Step 2: Choose Model ID

• Use the dropdown to select your preferred Model ID.

---

​

🌐 Step 3: Configure Language

• Input the 2-character language code that matches the target transcription language.
  • For example:
    • en for English
    • fr for French
    • es for Spanish

Refer to the documentation of the chosen provider (e.g., Deepgram) for supported language codes.

---

​

⚙️ Step 4: Utterance threshold MS

The Utterance threshold MS determines the threshold for the transcriber to consider the user’s speech as ended

• good range is between (150 - 500ms) anything less or more is not recommended.

---

​

✨ Step 5: Enable Input Voice Enhancer (optional)

• This will remove background noice, normalise and enhance input audio before transcribing it.

---

​

🚦 Advanced Settings

If your transcriber provider offers advanced options, such as custom models, input the Model ID field:

• Model ID: Use this if a new model has been released by the provider (e.g., Deepgram’s Beta models).
• Ensure to test thoroughly before deployment.

---

​

📝 Example Use Case

Here’s a simple scenario:

1. User: “What time is my meeting tomorrow?”
2. Transcriber: Converts voice into text: "What time is my meeting tomorrow?"
3. The AI processes the text and generates a response.

---

​

🔗 Next Steps

By configuring the transcriber settings, you can ensure fast, reliable, and accurate voice-to-text interactions in your