Emissions-TV - AI-Powered Contact Automation

CONTEXT & CHALLENGES

Teams faced with dozens of requests processed manually every day

Every day, Cassandra's teams manually handled dozens of requests (cancellations, modifications, recurring questions about shows, information requests during broadcast peaks…). Each request was processed manually by managers. This manual process was repetitive, time-consuming and unoptimized. The lack of weekend coverage (from Friday 7 PM to Monday 10 AM) turned Monday mornings into a high-pressure catch-up session.

The ambition was clear: reduce time spent on low-value tasks and the stress caused by Monday catch-ups and audience peaks, thereby improving manager satisfaction, streamlining the user experience and reducing response times on the user side, and aggregating and analyzing data on the organization side to use it to inform the user.

OBJECTIVES

Proof of Concept

Introduce a standalone AI-assisted contact processing module, integrated into the existing EMISSIONS TV platform. It automates request qualification, context assembly and response pre-drafting, while maintaining full manager control for validation and action.

The module was designed to operate 24/7, pre-process requests even outside business hours and transform the managers' workflow: from manual investigation to a fast review and validation approach. The strategic objective was to validate the approach and ROI under real-world conditions.

Measurable ROI: Hours saved x hourly cost = documented monthly savings.

• AI Agents
• NLP
• Automation

SOLUTION

A human-in-the-loop AI module