Lights, camera, action! How kinoheld transforms with Conversational AI

kinoheld  is a leading ticketing platform for movie theaters in Germany. Through its online portal, moviegoers can conveniently buy and reserve tickets for more than 500 theaters across the country. 

Challenge

When the latest blockbuster premieres, or holiday breaks are on the horizon, movie ticket purchases skyrocket. For kinoheld, these peak moments were leading to a maxed out call center and long wait times for customers. While initially, a rule-based AI assistant was deployed to fix the problem, it became clear that callers’ inquiries were too diverse and unpredictable to be handled by a purely rule-based agent. 

Solution

To reliably handle this diversity, the company’s automated intent recognition was switched to agentic AI. The AI agent not only better understands what customers mean with their questions, but it can also respond more flexibly. The agent is available 24/7/365 to answer questions about movies and the theater schedules. It helps customers make reservations by asking them to select their theater, movie, and showtime over the phone, then following up with a link to complete the booking via text message. The agent can also cancel tickets for customers who have already booked, diminishing the workload for both customers and call center employees.

Implementation

By working with Parloa and MUUUH!, kinoheld developed a 24/7 AI agent that steps in exactly where service bottlenecks occur: on the phone. The sheer volume and variety of information posed the biggest challenge for implementation. The AI agent has to be able to recognize the over 500 cinemas in the kinoheld network and correctly understand the corresponding movie titles. Now, the AI agent can accurately match movie descriptions and theater names, correctly identifying “the movie with the dinosaurs” as “Jurassic World” or “the new movie by Bully” as “Das Kanu des Manitu,” for example.

Through an API sync, the agent is regularly updated with data on topics such as accessibility, promotional offers, and theater amenities so that the agent can reliably handle all FAQ-related questions.

Results

A key success factor in implementation was the continuous, data-driven refinement of the AI agent. The transition from the rule-based solution to agentic AI proved measurable results:

• After just the first year, the agent was able to handle 50% of inquiries independently using only a single phone number. Today, this rate stands at 65%. 

• The recognition rate for cinema names rose to 85%.

• The average call duration decreased by 30 seconds. 

• 90% of FAQ-related questions are answered immediately with automation. 

“With the AI Agent, we’re improving accessibility for our moviegoers while ensuring our future scalability. It’s proof of successful innovation in service, and we believe it’s a solution that strengthens kinoheld in the long term.”

Peter Bedegi, IT Team Leader at kinoheld GmbH