Modern call-centers providers need state of the art quality controll and business analytics solutions. CLEMVOICE is a straightforward call-center analytics solution jointly developed by SpeechTex and Clementine Consulting.

Hungarian Bank

The behavior and speech of the customer service agents influence the customer satisfaction largely, therefore it is extremely important to monitor the protocol conformance, and to properly schedule and plan the needed trainings.

The system’s main aim is to make the call center performance transparent and make the evaluation of the agents automatically. Using CLEMVOICE the objective aspects of the evaluation growed from 1% to 100%, so the agent’s performance supervising is not based on few random calls but on all of them. After the introduction of the system not only the evaluation procedure became more efficient but the time used for the supervising decreased also, because checking of the calls became a data driven procedure (the QA staff listen the faulty conversations reported by CLEMVOICE), furthermore the QA staff do not have to check the aspects that the system has checked before.

CLEMVOICE supervise various protocol patterns, which are summarized and tracked at all conversation:  greetings protocol, identification protocol, negative attitude, prohibited words, break management and pause protocol, office routine protocol (based on the required script), telesales protocol. 

E-mail processing workflow is also accelerated by CLEMVOICE. If an e-mail goes to the adequate agent, the time spent on the processing is decreasing.The e-mails could be arranged in a priority order based on their contents, e.g. claims, urgent, forward, etc. With this proper ranking the difficult e-mails could be directed to experienced agents, and the simple could be managed faster, optimizing the answering time. Based on our simulation the system reduced the answering time of the claims with 20%.

Hungarian Insurance Company

The CLEMVOICE system has been working for two years at this company monitoring continuously the topic and the sentiment of the conversations, and detecting if any churn intention occurs.

On average 6000 calls per day are transcribed and analyzed by the analytic module resulting an overall rating of the call center and the agents. Our system rates each conversation, and gives daily, weekly, and monthly reports to support the supervising. Most crucial i.e., unsatisfied and churn risk calls are listed daily. Results of this quality monitoring can be integrated in the daily workflow aiming customer satisfaction and reduction of churn.

The CLEMVOICE system measures the performance of the agents based on the required protocol conformance. In each phone call the needed pattern’s appearence is checked:

- Identification protocol: the administration routine could start only after the verification of the required personal data (e.g. name, address, mother’s name etc.)

- Data collection: agents are monitored, whether they settled the asked data appropriately in the business data system (e.g. e-mail address, mobile number, willingness to receive marketing calls, etc.)

Based on the above mentioned measurements the system calculate a so-called protocol conformance score for all agent and call, in this way even the personal performance, even the entire staff’s performance could be tracked continuously. This is the basis of the evaluation and the training of agents.

Hungarian Telecommunication Service providers

As a result of fruitful R&D projects first call-center speech recognition implementations were demonstrated at telecommunication service providers. Our goal was to build up a flexible platform providing a basis for value-added services like indexing, topic identification, voice routing or business analytics. SpeechTex's results proved to be a useful and effective solution on this market.