"Saxony-Anhalt Digital 2030" is the strategy for a strong digital economy and modern administration in our federal state. Artificial intelligence is an important part of this, but it also poses challenges, especially with regard to the protection of personal data and control over where and how it is used.
We at Flying Circus now offer a solution for Saxony-Anhalt: an AI platform from the region, with full transparency, certified security and our high data protection standards. This allows us to leverage the benefits of AI-supported processes without relinquishing data sovereignty.
Our goal is to strengthen the digital self-determination of companies and public authorities in the state – in other words, the ability to use data and technologies in accordance with European Union legislation and as we all want, and ultimately to be able to control this through audits and compliance. With our platform, we are laying the foundation for new, innovative AI applications in Saxony-Anhalt that are one thing above all else: digitally sovereign!
Submit your project idea now!
This project is funded by the Ministry of Infrastructure and Digital Affairs, Department of Digital Strategy and Digital Projects. It demonstrates the state's commitment to future-oriented and secure digitalisation. Are you interested in realising your project on this platform? Let's discuss your project or idea in a short video meeting.
Solution 1: Our AI platform
- • Fast implementation and realisation time
- • Own database (RAG) for proprietary AI knowledge
- • Multiple LLMs for different applications to choose from
- • High data protection and ISO certified
Solution 2: The custom AI project
- • Highly customisable AI and ML workflows possible
- • Dedicated hardware on request
- • High power and performance
- • Flat-rate billing models available
Our AI solutions in detail:
Our solution is based on a proven technology stack that combines flexibility and security. At its core is a secure data hub with a powerful vector database that not only forms the basis for Retrieval Augmented Generation (RAG), but also ensures the reliable storage and processing of your data. Depending on your use case and resources, you can choose from a range of large language models (LLMs) designed for optimal performance and accuracy. Integration into your existing IT landscape is simplified by open API interfaces and a user-friendly web UI. Access to data and functions is protected by comprehensive user management and strict access controls.
Our experienced team develops customised applications for you that are tailored precisely to your needs – from conception to implementation and maintenance. This means that we not only adapt existing modules, but also design and implement complete solutions. The spectrum ranges from direct GPU control for particularly computationally intensive tasks – for example, real-time processing of large image data sets in quality control – to the intelligent integration of different technologies. From text recognition in image data, classic machine learning algorithms for process optimisation, or responsive chatbots with RAG integration. No matter how complex your requirements are, we will develop a solution that is perfectly tailored to your challenges.
RAG included – More knowledge for your AI solution
Having AI-powered processes is good. Having AI that is up to date and provides informed answers is even better. Retrieval Augmented Generation (RAG) is a technology that does just that. RAG extends the capabilities of LLMs by giving them access to specific, relevant knowledge – especially your own proprietary knowledge.
Björn Langer
Let's talk about your project!
As your first technical contact, I am committed to understanding your individual requirements and developing tailor-made approaches for the successful operation of your application. Our exchange also enables us to determine an initial price indication for your individual project. I look forward to talking to you!
Contact options
How does RAG work?
RAG combines the strengths of large language models with the precision of a targeted knowledge database, known as vector databases. The process can be divided into three steps:
- Knowledge preparation: From text to vectors: Your documents, reports and data are converted into a form that an LLM can process quickly and easily. The text content is converted into numerical representations, known as vectors. These vectors capture the meaning of the content and enable the AI to quickly identify similarities between different texts.
- Finding relevant information: When a question or query is received, the AI does not search the entire internet for answers. Instead, it searches the prepared knowledge database for documents that are most similar to the query. Thanks to vector-based search, it can find the relevant knowledge quickly and accurately.
- Answers with substance: The documents found are then used to answer the original query. The AI not only draws on its general knowledge, but also combines this with the specific knowledge from your documents and provides an important source reference. This results in well-founded, accurate and context-related answers.
Why is RAG so valuable for your AI solution?
Modern large language models are impressive, but they also have their limitations. They often lack access to current, specific information or are unable to cite sources for their answers. RAG solves these problems:
- Always up to date: RAG enables your AI solution to always respond with the latest knowledge without having to retrain the entire model.
- Transparency & traceability: RAG provides answers based on specific sources within your data. This allows you to verify the accuracy and relevance of the information.
- Reduction of inaccuracies: By limiting the AI to your own verified knowledge, RAG significantly reduces the risk of hallucinations and the spread of false information.
- Protection of sensitive data: RAG minimises the risk of AI reproducing sensitive or confidential information from its original training. With RAG in our platform, we provide the foundation for your modern AI application that not only has your specific knowledge, but is also reliable, easy to update and compliant with data protection regulations.
Data protection compliance as a foundation – AI use in compliance with the BDSG and GDPR.
In times of strict data protection regulations, compliance with the BDSG and GDPR is essential for companies and public institutions. Our AI platform has been designed from the ground up to meet the highest data protection standards.
Our commitments:
- Data processing exclusively in Germany: Your data will be processed and stored exclusively on servers in Germany.
- Data processing agreement (DPA): We offer a comprehensive DPA that meets all relevant GDPR requirements.
- Binding data deletion periods: We guarantee compliance with binding data deletion periods.
- Transparency and traceability: You have access to your data at all times and the right to audit our solution.
- ISO-certified processes: Our platform, all applications running on it and our primary data centre in Oberhausen are certified according to the international standard for IT security ISO 27001.
Contact us today for a no-obligation consultation! Take advantage of the funding opportunity offered by the state of Saxony-Anhalt now! We offer you the opportunity to reduce the initial costs for your AI project. The only condition: Your project should have a direct connection to our state – either regionally through its location or digitally through its target group or area of application. Together, we will develop a tailor-made AI solution that meets your requirements.