<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=2171572209666742&amp;ev=PageView&amp;noscript=1">
conversational_ai_callcenter_header

AI Co-Pilot at CSS Insurance

Agents in the CSS Insurance Service Center now use an AI Co-Pilot to find information quickly.
https://www.adnovum.com/hubfs/FILES/06_Careers/Contact%20People/sascha_kretz.jpg

«The Co-Pilot makes our customer service more attractive for both customers and employees.»

Sascha Kretz

Head Service Center, CSS Group

The starting point

Every year, the CSS Service Center handles over 1.7 million telephone calls and numerous written inquiries such as e-mails. The employees process them manually. Previously, they had to call up several peripheral systems by hand, such as the website, intranet, and help documents, in order to find a detailed answer. This took a lot of time and made building up knowledge difficult, in particular for new employees, who are increasingly called in to help out during peak periods such as the fall season.

Facts and figures

  • 400+

    Agents

  • Trilingual

    GE / FR / IT

  • CRM

    Seamless integration

insurance_new_business_3

Features that give real added value

The seamless integration of Co-Pilot relieves agents directly in their familiar working environment.

Faster access to information in real time shortens customer consultations.

The answers can be given and used more quickly and reliably.

Callers are more often able to resolve their concerns on the first call, leading to greater customer satisfaction.

The context of questions earlier in the interview is retained and taken into account when generating the answers.

The quality of answers is continuously improved by evaluating the replies of the employees (up-/downvote).

How we got there

Together with Adnovum, CSS aims to use artificial intelligence to assist agents in their work: with a new, specially designed Co-Pilot. This AI assistant searches for sources in peripheral systems, such as the website, the intranet and numerous documents, to provide in particular new employees with the necessary knowledge and to help them formulate accurate responses. The procedure was as follows:

CSS and Adnovum have implemented a Co-Pilot based on a retrieval-augmented generation (RAG) solution to combine the benefits of semantic search and large language models. This project is part of the multi-year conversational AI roadmap of CSS, which aims to make customer service increasingly attractive for customers and employees alike by means of AI. Adnovum supports CSS with a multidisciplinary team of conversational AI experts – from conception to implementation to operation.

The Co-Pilot was seamlessly integrated into the CRM solution of the CSS Service Center. Agents can now ask questions about CSS products, processes, and services via chat, and the Co-Pilot provides consolidated answers, including links to their sources. This allows Service Center employees to view the original document if required and check the information in its original context.

«We are glad to work with Adnovum on our conversational AI roadmap for a quick and beneficial introduction of modern solutions and to make the most of the novel opportunities of generative AI.»

Sascha Kretz

Head Service Center, CSS Group

Technical implementation

css-rag-setup_en

The architecture was implemented with a cloud-native approach on Microsoft Azure and rolled out automatically with Terraform. The solution automates the regular updating of the knowledge database (vector database) and obtains information from external and internal websites, the contact center intranet, and PDF documents. With Microsoft Prompt Flow, the entire prompt for the request can be iteratively adapted to the OpenAI model GPT-4o. Above you can see an overview of the basic setup, which is now being expanded continuously.

This is our client

The CSS Group in Lucerne is one of the largest health insurers in Switzerland with over 1.6 million insured persons. The company plays a central role in the Swiss healthcare system by providing access to high-quality medical care and strengthening solidarity in health services. Through prevention programs, CSS promotes the well-being of the population and contributes to the long-term reduction of medical costs. This commitment of CSS underlines its importance as a socially responsible company.