Smart summary

Updated at:

RTC smart summary, powered by the Qwen Large Language Model (LLM), provides a new way to record and process audio and video content. This topic describes what smart summary is and how to use it to meet your business needs.

Concepts

At its core, RTC smart summary is a speech-to-text service that converts spoken content into text and stores the transcripts in your specified Object Storage Service bucket.

Scenarios

RTC smart summary supports the following scenarios:

  1. Enterprise office: Integrate RTC into office systems such as Office Automation (OA) and Customer Relationship Management (CRM) to improve information extraction efficiency for meetings, interviews, training sessions, and customer communications. You can trace conversations, query key points, and quickly locate relevant content.

  2. Online education: Summarize key points, divide lectures by subject, and add captions to online courses. This helps you maximize the value of lecture videos and makes review more focused.

  3. Finance and media: Transcribe and summarize content from interviews, speeches, and roadshows.

  4. Sales and customer service: Convert conversation audio to text for service quality inspection. Extract important information such as customer needs, purchase intentions, and feedback to increase sales conversion rates.

Features

  1. Speech-to-text transcription

  2. Chapter overview

  3. Summarization (full-text summary, speaker summary, Q&A review)

  4. Key point extraction (to-do items, keywords, highlights)

  5. Spoken-to-written language conversion

  6. Service quality inspection

  7. Custom prompts

Steps

To improve speech recognition accuracy, integrate the intelligent noise reduction module of RTC on the client. This produces more accurate transcripts.

  1. Integrate intelligent noise reduction (recommended). For more information, see Intelligent noise reduction.

  2. Join a meeting using the RTC software development kit (SDK). For more information, see SDK download.

  3. Start a summary task. For more information, see Start a summary task.

  4. Stop a summary task. For more information, see Stop a summary task.

  5. Receive summary callbacks. For more information, see 3001 Summary successful.

  6. Query summary results. For more information, see Summary file list.