AI in practice: the Winweb chatbot

Gen AI, short for Generative Artificial Intelligence, is a term that is currently on everyone's lips in the tech world. This technology marks a turning point in the way we understand and use digital assistance. A prime example of the practical use of Gen AI is ChatGPT, which has already been integrated into the daily lives of many people. Winweb recognized the potential for its software solution winweb food and quickly sought the right partners to enable its own customers access to this technology. From now on, each of our customers will have an advanced chatbot at their side. This intelligent assistant is trained to answer questions about Winweb by drawing information from the WinwebWiki and the winweb food help.

Thanks to the successful collaboration with our partner DataStax we have implemented a chatbot based on RAG technology (Retrieval-Augmented Generation). This enables our AI to generate answers to complex queries through targeted searches across a wide range of information sources, resulting in significantly greater relevance and precision of the information provided. The potential of this technology is enormous and opens up new horizons for interaction between customers and our service.

First, content from various sources, e.g. our WinwebWiki and the winweb-food help, is converted into so-called embeddings. This technical method makes it possible to convert text into dense vectors, which are then stored in a specialized Datastax Astra vector database. AstraDB enables complex, context-sensitive search queries in large data sets and stands out through the ability to deliver scalability and performance for generative AI applications without developers having to fear high operational overhead. Access to this database takes place via a customized service that was developed with Python, Flask and Docker and can scale as a cloud service depending on the volume of requests. This service serves as an interface that receives requests from Winweb apps or from winweb-food, searches the corresponding embeddings in the vector database and returns the most relevant documents in real time. The artificial intelligence then has access to all relevant documents that are useful for answering the user request.

Winweb and DataStax are committed to the highest data protection standards and ensure the security of customer data through current encryption standards and the operation of all servers within the European Union. This step ensures that all data processing activities are subject to the strict EU data protection guidelines, which are among the most robust worldwide.

We warmly invite you to take part in this exciting journey. Are there specific data sources or systems in your company that you would like to connect to your personal chatbot? Do you have special requirements or requests? We are open to your ideas and suggestions in order to continue developing our service together.

Contact us directly to learn more about how Winweb and our innovative AI approaches can support your company. Let us push the boundaries of what is possible together.