From the past, the quality inspection department of the contact center has been conducting random inspection on the recordings manually. With the rapid development of full media contact centers, the business volume has increased dramatically. Hence, traditional quality inspection can no longer comply the requirement of modern contact centers. Therefore, it is necessary to carry out quality inspection efficiently and accurately, which can only be realized by relying on advanced technology.
 
 

Current Situation of Manual Quality Inspection


   
Low Coverage   Low Business Value   Delayed Quality Inspection
The random inspection in current industry is around 1-2%, which means that a large volume of recordings are ignored. As a result, hidden values or risks have not been effectively probed.   Due to the low coverage rate, manual quality inspection cannot analyze all the recordings and accurately predict the customer need to secure business opportunities. Hence, the business value of quality inspection personnel cannot be well reflected.   Manual quality inspection is generally a random inspection after an event. This method cannot probe the problem promptly and unable to respond to risk in time.
         
   
Low Working Efficiency   Unstandardized Quality Inspection Standard    High Quality Inspection Cost
For each randomly selected call, quality inspectors often need to listen to it repeatedly. This results in low efficiency and only limited problems can be found during working hours.   Random inspection manually is limited by the cognitive of business. Different people often have different perspective about event judgement. Even the one’s perspective on the same event will change at different times, which results in the inconsistency of standards.   Quality inspection is a high repeatability and heavy task. With the increase of business volume, the company is required to invest more on manpower and material resources to meet the needs of quality inspection and ensure a great proportion of random inspection.
 
 

VoiceCyber Intelligent Quality Inspection Solution


“Intelligent Quality Inspection” is realizing transcoding, analysis and data mining on voice data based on speech recognition, speech analysis and big data processing technology. This provides the required resources and application service of comprehensive quality inspection management system for data processing. It ultimately assists in compliance inspection, risk analysis, trend analysis and business opportunities probing by analysis and mining from the recorded data via a flexible quality inspection module.
 
 

Functional Characteristics


Speech Recognition

By using advanced speech recognition technology, voice recording files are analyzed and processed through voice transcription, speaker separation, mute detection, speech rate detection, emotion detection and keyword extraction. The purpose is to generate corresponding text analysis results, which provides support for quick searching, listening and analysis of the long duration recordings.

 
Full Text Search

 

The system provides rich and flexible query conditions for full-text searching, which can highlight the searching keywords and clearly mark the conversation content. The colour-based separated incoming and outgoing conversation can be directly and positioned for playback by just clicking on key sentences.

 

Visualised Playback

Support multi-speed playback, text following and positioned playback.
Visualised display of voice, content and timeline. Clear separation of incoming and outgoing conversation.

 
Quality Inspection Module

 

Customize rules according to business process with full data coverage and inspection.
Flexible quality inspection module, including quality inspection scenarios such as service standards, polite language, use of opening and closing words, speech rate, silence time, etc. The search results can be quickly located and playback.

 

Fully Automatic Quality Inspection

The ability and service attitude of agents are automatically scored according to the configured quality inspection rules (including keywords, emotional index, mute index, etc.). High coverage is ultimately achieved and the problem of low coverage by manual quality inspection is solved.

 
Manual Evaluation

 

The system will screen the calls with low scores for the quality inspectors to listen. This forms a new quality inspection mode “intelligent quality inspection and extensive listening + intensive listening by quality inspectors”. The final goal is to comprehensively grasp and improve the service quality of agents in customer service or contact centers.

 

Data Mining

Using systematic method to mine the value of massive voice data in storage and avoid the loss of useful information.
Perform visualised display of data analysis, call statistics, hot spot distribution and anomaly analysis on the voice of quality inspection. The reports provide a reference basis for the management in decision making.