Intelligent Contact Center provides a communication AI engine, communication agents, manual agents, assistant agents, and intelligent outbound robots. You can select different features based on your business needs.
Communication AI engine
The communication AI engine integrates speech recognition (ASR) and text-to-speech (TTS) to interact with large models by streaming text. It detects and controls events such as interruptions, silence, and barge-in during calls to create a natural, human-like conversational experience.
Feature set | Feature | Description |
Communication AI engine | Create a large model gateway and test its availability to enable association with large model applications. Note You can use large models from Alibaba Cloud Model Studio (Bailian), third-party platforms, or your own models. To ensure compatibility, models must comply with the communication engine gateway protocol. For more information, see Large Model Gateway Integration Parameter Protocol. | |
Associate a large model gateway and configure personalization settings and call event handling. | ||
Improve speech recognition (ASR) accuracy by adding difficult-to-recognize words to a hotword vocabulary. | ||
This feature generates a realistic, personalized voice by using a large model to analyze an individual's voice samples. The model simulates unique vocal characteristics, such as timbre, intonation, and rhythm, to make interactions feel more natural and authentic. | ||
Receipt message configuration | You can receive call records and recording messages by subscribing to Message Service (MNS) in consumer mode or by using the HTTP batch push mode. |
Communication agent
A communication agent is an out-of-the-box, large-model-powered voice calling solution that replaces manual agents. It handles calls and engages customers in natural, fluid conversations, enabling you to efficiently convert leads, send reminders, and provide customer service while significantly reducing labor costs.
Feature set | Feature | Description |
Communication agent | Create a knowledge base for the agent using business resources like private knowledge files, voice recordings, and ASR hotword vocabularies. | |
Define the background, objectives, requirements, and frequently used knowledge for a task, and associate it with a private knowledge base. You can also configure the call timbre and volume, define exception handling, and link a personalized timbre and hotword vocabulary. | ||
After verifying the agent's performance through test calls, launch an outbound task to achieve your business objectives. |
Intelligent outbound robot
This feature provides a comprehensive interface for designing robot scripts for various business scenarios. These scripts allow the robot to automatically contact end-users through predefined outbound tasks.
Feature | Description |
After configuring the script and recording the audio, submit it for review. Once approved, it can be used in outbound tasks. | |
You can copy a template to 'My Scripts', make minor adjustments if needed, and submit it for review. | |
Create and run outbound tasks, and view detailed execution reports. |
Assistant agent
Call summary: During a call, the agent dynamically identifies key information, such as customer intent and business needs. After the call, the system generates a concise, multi-tagged text summary, allowing you to quickly grasp the conversation's main points.
Call quality inspection: You can pre-configure quality inspection rules, such as standard service language and process integrity. The system then automatically scores calls and flags anomalies, which helps improve service quality, ensure compliance, and reduce operational risks.