Changelog

Fallback Voices in Agents

Ensure smooth and reliable voice calls with customizable backup voice options.

How It Works

Fallback Voices allow you to pre-select one or more backup voices when configuring an agent. If the primary voice fails during a call, the system automatically switches to the first available fallback voice. If no fallback voice is set, the system continues with the existing automatic selection behavior.

The UI clearly shows that fallback voices are optional, so you can configure them only when you want added reliability.

• Fallback Voices selection was added to the agent editor.
• Automatic failover switches to your chosen backup voice if the primary voice fails.
• Optional setup means nothing changes if you don’t configure fallback voices.

Why It Matters

• Improved reliability: Calls continue without interruption even if a primary voice fails.
• Consistent experience: Reduces the risk of jarring or unexpected agent voice changes.
• Greater control: Choose how your agents sound in different scenarios.
• Standards aligned: Matches Retell’s API capabilities for voice configuration.

This update is now live and available in the agent editor.

Real-Time Speech-to-Speech (s2s) Voice AI

Whippy now supports real-time speech-to-speech AI models, enabling natural, instant voice conversations powered by OpenAI’s realtime models.

How It Works

Speech-to-speech (s2s) models process audio input and generate audio output directly. Unlike traditional setups that rely on separate speech-to-text (STT) and text-to-speech (TTS) systems, these models handle everything in one step. This creates a faster, smoother, and more natural experience inside Whippy.

Whippy now supports three realtime models:

• gpt-4o-realtime – best for advanced, natural conversations
• gpt-realtime – balanced for general use
• gpt-4o-mini-realtime – lightweight, cost-efficient option

When setting up a voice agent, you can:

1. Select a realtime model based on performance vs. cost needs.

2. Pick a voice from OpenAI’s available options (male, female, neutral).

3. Start a live conversation directly in Whippy with real-time audio responses.

4. Reconfigure models or voices between sessions as needed.

Why It Matters

Low latency (~200ms): Conversations feel smooth and responsive.

Natural voices: High-quality, pre-trained options ensure human-like tone and flow.

Unified experience: A single model handles the process, reducing complexity and integration overhead.

This makes it easier to build real-time voice agents for sales calls, support lines, or interactive demos — all within Whippy, without relying on separate voice providers.

Assign AI Chat Agents from the UI

You can now manually assign, change, and unassign AI Chat Agents directly from the conversation interface.

How It Works

We added a new option in the conversation header and conversation detail view that lets you manage AI Chat Agents with a single click.

• Assign an Agent: Select the agent you want to assign and click Assign Agent.
• Change an Agent: If a conversation already has an active session, switching agents will end the current one and start a new session after confirmation.
• Unassign an Agent: Click Unassign Agent to end the current session.

Before this update, Chat Agents could only be triggered through automation rules in sequences, campaigns, or keywords, or by making calls to the public API. Now, users can launch and manage Chat Agents directly from the UI without depending on automation or code.

This change makes it easier for teams to decide when and how AI agents take over a conversation, giving more control to the user.

Why It Matters

With this update, you now have three ways to activate Chat Agents:

1. Automations in campaigns, sequences, or keyword triggers.

2. API calls.

3. Manual assignment from the UI.

This flexibility means you no longer need to rely solely on triggers or developers to start Chat Agent sessions. Teams can now offload conversation handling to AI agents whenever it makes sense—directly from the Whippy interface.

Start a Chat Agent via Public API

You can now start a chat agent using our Public API. This makes it possible to trigger agents directly from your own systems.

How It Works

We added a new Public API endpoint that starts a chat agent.

Previously, agents could only be initiated through triggers and actions in sequences, campaigns, or keywords. With this update, you can now call the endpoint directly.

Why It Matters

This change gives you more control over automation. You can trigger a chat agent based on events in your own system, not just within our platform’s workflows.