Voice Intelligence is now Conversational Intelligence. Learn more about our expansion to new channels.
Transform every interaction into valuable insights with Conversational Intelligence
Virtual agent and Messaging support
Conversational Intelligence now integrates with ConversationRelay virtual agents and Messaging channels to bring unified AI insights across customer touch points. Sign up for the Conversational Intelligence for Messaging beta.
How does Conversational Intelligence work?
Extract customer insights and business intelligence anywhere you interact with customers.
Channels
Agents
Capabilities
Uncover opportunities to drive operational efficiency and improve customer satisfaction through Conversational Intelligence.
Turn your conversations into actionable insights. Start for free.
Twilio Conversational Intelligence lets you act on conversational data with one convenient, AI-driven tool. View our pricing for Conversational Intelligence for Voice or sign up for free today.
FAQs
Conversational Intelligence is an AI-powered tool that analyzes voice and messaging conversations in real time. It’s designed to help businesses understand, automate, and improve customer interactions by transcribing calls, extracting insights, and enabling advanced features like virtual agents.
It works by:
- Call routing - connect the call to the Conversational Intelligence service
- Real-time streaming - call audio is streamed to the Conversational Intelligence service via websocket. Your application can interact with the stream to send and receipve messages like prompts, responses, and interruptions
- AI processing - AI transcribes the conversational as it happens. AI models (Called language operators) analyze the conversation for sentiment, task completion, virtual agent performance, interruption detection, human agent escalations.
- Transcripts and insights - a detailed transcript is generated with speaker labels and results of AI analysis.
You can use Conversational Intelligence for AI-powered call analysis and text messages analysis.
Voice
|
Source |
Status |
|---|---|
|
Twilio Recordings |
Available |
|
External Recordings |
Available |
|
Calls |
Available |
|
New: ConversationRelay |
Available |
Messaging
|
Source |
Status |
|---|---|
|
New: Conversations |
Language Operators are modularized Natural Language Understanding (NLU) capabilities, powered by AI. Language Operators analyze your customer conversations and turn the raw unstructured data from voice calls and messages into structured information that can be used to generate transcripts, analyze interactions, detect sentiment, and monitor agent performance.
WIth Twilio, you can choose to use pre-built language operators for out-of-the-box functionality like sentiment detection and call recording disclosures. Or, you can create your own custom language operators for your specific analysis use cases. The inferences from the operators will show up in the conversation transcript.
Yes, Conversational Intelligence provides a single unified platform to monitor the performance of both human and virtual agents. This combined visibility gives you insights to understand a customer’s end-to-end journey.
You can set up Language Operators that are used across both agent types (human and virtual), or specify certain Operators that work specifically for human vs. virtual agents.
Twilio Conversational Intelligence provides industry-leading accuracy. You get the choice of best-in-class speech-to-text providers, with each engine and model thoroughly vetted by our in-house Machine Learning (ML) team. Conversational Intelligence transcription is also natively integrated with other Twilio communications channels, so you have a simple and seamless way to turn any communication media into a transcript that can be analyzed.
Twilio Conversational Intelligence supports both real-time and post-conversation transcription for voice calls. Depending on your use case, you can select which makes the most sense for you.
The platform currently supports 11 languages:
- English (US, UK, AU)
- German (DE)
- French (FR)
- Italian (IT)
- Spanish (MX, ES, US)
- Portuguese (BR, PT)
- Dutch (NL)
- Norwegian (NO)
- Polish (PL)
- Swedish (SE)
- Danish (DK)
Users should ensure compatibility with their preferred language before implementing the platform.
Yes, Twilio Conversational Intelligence is provider agnostic. You can use any stereo/dual channel recording and convert it for action. We believe you shouldn’t be boxed in to one provider and want you to have the freedom to integrate Twilio with your existing tech stack.
Dual channel recording: You can separate each speaker onto individual audio channels to improve speech separation, transcription accuracy, and speaker identification.This ensures clearer recordings, more precise sentiment analysis, and better keyword spotting, making it an essential component for delivering reliable and insightful voice analytics.
Customer sentiment analysis: Sentiment analysis helps you extract valuable insights from voice calls and messages. Twilio Conversational Intelligence acts like a keen listener, going beyond words to detect emotions and determine if sentiments are positive, negative, or mixed. This deeper understanding of customer feelings helps businesses track sentiment trends and respond effectively, ensuring their actions resonate with customer needs.
Call summarization: Conversational Intelligence takes long chats and boils them down to the essentials. It gives you the key points—like what was talked about, what needs to be done, and any big decisions made—so you can focus on what matters, make decisions faster, and improve your customer experience.
Both conversational intelligence and call tracking involve analyzing customer interactions, but they serve different purposes. Call tracking focuses on capturing call data for marketing attribution, whereas conversational intelligence provides deeper insights into the content and context of conversations, enabling more strategic decision-making.
Call recording is… but Conversational Intelligence applies AI to analyze the conversations, so you get insights at scale.