Bouygues Telecom and IBM: Concisely summarising customer calls

Challenge: As digital transformation accelerates, telecoms operators urgently need to automate the production of call summaries and related recommendations. Although the millions of conversations between customers and agents can generate many insights, agents can only capture part of this valuable information in the CRM (customer relationship management) systems. When there aren’t sufficient details in the CRM system, agents don’t have the time to fully read automated transcriptions of calls.

Solution: Working with IBM, Bouygues Telecom is using AI to create concise summaries of customer-agent conversations and update its CRM system to facilitate quick reviews of previous calls. The system also extracts actions to ensure consistency between documented actions and the call transcript. At the same time, cross-referencing mentions of the competition and sentiment analysis helps flag and understand the risks of customer churn. These improvements are designed to enhance operational efficiency and effectiveness.

Impact: Bouygues Telecom has identified the potential to make operational improvements over three months valued at more than US$5 million. The system is also set to reduce pre call and post call operations by 30%.

Topic: Customer experience, and language and content AI (NLP/LLM/GenAI)

Cost: Medium

MNOs and supporting organisations: IBM and Bouygues Telecom