MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641051197 A) filed by The Principal, Sns College Of Technology on April 22, 2026, for System And Method For Ai Driven Voice Based Customer Feedback Analysis In Call Center Recordings.
Inventors include Jagadeesh. B; Jeya Chandran K; Rudhuvarshan. R; Akshay Chandra. N; and Raahin Suhail. S.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: An Al-driven voice-based customer feedback analysis system and method for call center recordings is disclosed. The invention addresses the critical limitation that only 2 to 5 percent of call recordings are currently reviewed manually, leaving the majority of customer feedback unanalyzed. The system employs a multi-stage processing pipeline comprising a Speech Processing Unit (.104) with a transformer-based Speech-to-Text Engine (104a) and Speaker Diarization Module (104b) that transcribes and separates customer and agent utterances from audio recordings. An Analysis Engine (106) processes diarized transcripts through three parallel Al modules: a Sentiment Analysis Module (106a) that classifies emotional polarity at utterance, call, agent, and campaign levels; an Emotion Detection Module (106b) that identifies discrete emotional states including frustration, satisfaction, confusion, and anger to enable real-time escalation detection; and a Topic Extraction Module (106c) that employs Named Entity Recognition and semantic clustering to categorize call content into actionable business topics. A Confidence Scoring Layer (108) applies configurable threshold logic to validate analytical outputs, routing low-confidence predictions to secondary validation while passing high- reliability insights to a Persistence Layer (110) supporting both historical analytics and real-time alerting. A Dashboard Interface (112) delivers aggregated sentiment trends, emotion timelines, topic heatmaps, and agent performance scorecards to Customer Experience Managers and Call Center Supervisors through a role-based, web-accessible visualization layer. The invention provides enterprises in telecom, banking, insurance, e- commerce, travel, and healthcare sectors with a scalable, automated solution for transforming raw call recordings into actionable customer experience intelligence.
Disclaimer: Curated by HT Syndication.
