Voice remains the dominant communication channel for customer service. In insurance, callers are stressed, and the information they give is messy and out of order. Call center agents on the other end need to navigate complex conditional question trees while trying to capture and structure the information in real-time.
At QuantCo, we have built LLM-based systems covering voice channels. A voice assistant sits alongside call center agents and supports the call by proposing relevant questions and extracting structured data. At peak times, our voice agents can go further: a low-latency loop handles calls end-to-end autonomously.
In the session, we will walk through the design decisions to satisfy the constraints in latency, quality, and cost. Together, we will try out real-case scenarios by live-calling our AI system.
The workshop itself is open to all students and does not require an application.
The presentation will start at 6 pm on May 28th. Afterwards, we will invite a selected group of students to a sit-down dinner in Aachen with several QuantCos. Applications are only required for the dinner.
Feel free to reach out to us: events@data-science-club.de
