National Sun Yat-sen University / Prof. Chien-Cheng Wu 

 Pain Points Solved 

This technology provides enterprises with an in-house call recording system and subsequent intelligent management mechanisms, enabling them to comprehensively retain voice records of transactions and service interactions as a basis for post-event verification and management. The pain points it addresses include: 

• Reducing the risk of transaction or service disputes: By preserving complete recordings of transaction and service interactions, the system provides supporting evidence when disputes arise, thereby enhancing the enterprise’s ability to protect itself and clarify responsibilities. 
• Supporting call center quality inspection needs: It helps call centers establish standardized recording management and sampling review mechanisms, improving the efficiency of service quality monitoring while also serving as a foundation for staff training and performance management. 
• Capturing the voice of the customer to support operational decision-making: Through the collection and analysis of voice data, the system helps enterprises understand customer needs, frequently asked questions, and service pain points, thereby supporting product improvement, process optimization, and operational strategy analysis. 

 Technology Introduction 

This platform provides enterprises with an in-house call recording system and intelligent speech analytics services, transforming recordings from transactions, customer service, and other service interactions into manageable, searchable, and analyzable data assets. The platform features speech-to-text conversion, keyword detection, and customer insight extraction, helping organizations reduce transaction and service disputes, improve call center efficiency, and derive customer needs and market trends from large volumes of voice data. These insights can further support operational strategy adjustments and service process optimization. Combining practical deployability with strong commercial value, the platform is well suited to serve as a key tool for enterprise digital transformation and intelligent management. 

▲Figure. System Architecture Diagram: Intelligent Speech Audit and Conversation Quality Analysis Platform (ASCAP) - Main Functional Blocks

 Application Examples 

This platform addresses the limitations of traditional manual quality sampling, where only a very small portion of recordings is typically reviewed, often just 1–2%. By leveraging AI-based automatic speech recognition (ASR) and natural language processing (NLP), it enables 100% full-coverage auditing of recorded interactions. 

• Standardized recording management: The platform automatically consolidates conversation records across multiple channels, including phone calls, online chat, and mobile apps, creating a unified digital archive with searchable indexing.
•  Automated screening and review: Built-in quality inspection rules, 
  such as greeting compliance, polite language usage, and prohibited phrase detection, allow the system to automatically identify high-risk or abnormal conversations, so human reviewers can focus their efforts on cases that truly require intervention.
• Performance evaluation and training support: Through data visualization and performance analytics, the platform helps identify knowledge gaps and service weaknesses within customer service teams, providing an evidence-based foundation for developing training materials and continuously refining standard operating procedures (SOPs). 

 Related Links 

None

 Patent Name and Number 

None

 Industry-Academia / Tech Transfer Partner 

None

 Honors and Awards  

None

 Technical Contact  

Industrial Liaison Office

National Sun Yat-sen University
Tel: +886 7-5250165
Email: gloria@mail.nsysu.edu.tw