Lacking clarity in customer sentiment and spending USD 2 million on tools for quantitative analysis, a worldwide home appliance, and vehicle components seller turned to Infosys for an innovative solution.
Driven by the need to better understand their customers, the company sought to significantly enhance their analysis of customer-agent communications. Their goal was to capture more nuanced insights from the tone and sentiment of these dialogues, while also seeking substantial cost savings.
Key Challenges
- Realize greater depth and accuracy in the analysis of consumer-agent communications (emphasizing the tone and sentiment of both speakers)
- Drive down costs connected to this analysis
Ready to experience?
TALK TO EXPERTS
The Solution
Using AI to generate insights from voice-based conversations
Infosys delivered a solution using Infosys Topaz and partnered with Google, using tools such as Vertex AI Text Bison foundation and Chirp. These tools could be deployed to deliver text summarization of voice conversations between agents and consumers – identifying the category or product being discussed in these conversations and then generating insights from them. The outputs could be delivered in 2-3 days, whereas a month was needed previously.
Transforming customer sentiment analysis through AI
- Quick, accurate text summarization of voice conversations
- Identification and analysis of discussed product categories
Benefits
Lower costs: Realized a 50 percent reduction in its data analytics costs
Greater accuracy: Achieved a 25 percent improvement in transcription accuracy
More precision: Captured more nuanced emotions and aspect-oriented analysis than previous models
