VIER AI Gateway: Use AI now in a legally compliant, simple and transparent way

For many companies in Germany, the use of generative AI is already an excellent way to automate processes, relieve the burden on sales, marketing and service through bots or speed up document processing. So far, however, the data used has often been openly transferred to the generative AI or its manufacturers (for example in the USA). Even with guardrails, i.e. the guardrails that restrict the use of an AI solution within certain specifications, companies are dependent on the manufacturers and cannot introduce their own specifications or restrictions. It is also difficult to understand, let alone prove, which data is used and what costs are incurred by which bot deployment.
“However, the introduction of the AI regulation is now increasing the pressure to ensure secure, verifiable and transparent AI usage, in which data is protected and usage is logged,” explains VIER CEO Rainer Holler. “With VIER AI Gateway, this is now possible - simply and easily!”

Extensive functions make legally compliant, cost-optimized AI easy

VIER AI Gateway offers extensive options for the legally compliant use of AI. It is a kind of middleware between a wide variety of business applications, e.g. email automation, chatbots, voicebots, automated document creation/editing, social media apps or other customer-specific applications, and the various generative AI models. The following functions are available:

• A unified API enables access to the various AI models via an API that is compatible with that of the market leader OpenAI. This means that any application that previously accessed the OpenAI models directly can be very easily converted to secure use via VIER AI Gateway. Companies can also switch from OpenAI to other models such as Gemini from Google without making any changes to the application.

• AI Selector regulates the selection of the most suitable AI model for the use case and enables cost-optimized model usage. VIER can also recommend suitable models for different applications, drawing on the knowledge of its own research department.

• AI Profiles is used for profile management. Companies can set up AI profiles to control access to AI models for departments, users or applications. This allows costs to be allocated precisely and usage to be documented.

• Logging/Audittrail takes care of detailed activity logging. This ensures traceability for companies, as required by the EU AI Act.

• Dashboard ensures transparency. It provides an overview of the use of the various AI models and their assignment to different departments and applications within the company.

• Privacy Manager protects personal and other relevant information from companies that should not or may not be transferred to an AI, such as customer names, dates of birth, etc. To process requests containing such data, the data is anonymized or pseudonymized before it is transferred to a generative AI.

• Guardrails are the central tool for describing and enforcing guidelines that a company defines or must adhere to when using AI models. The guardrails increase the reliability of answers and prevent unwanted statements and hallucinations.

• Analytics enables comprehensive analyses of AI usage in the various applications by different areas of the company.

Stress-free AI process automation without risks

“We are incredibly proud that with VIER AI Gateway we are providing a comprehensive, yet easy-to-use solution for legally compliant and transparent AI usage that is second to none!” comments VIER CEO Rainer Holler on the commercial launch of VIER AI Gateway. “This gives companies of all sizes and from all industries the chance to really benefit from the enormous possibilities of process automation through AI without taking any risks,” says Holler. “VIER AI Gateway offers all the relevant options in terms of legal security, data protection, transparency and traceability that AI manufacturers do not offer. This brings enormous benefits for companies and their customers. And it massively increases the attractiveness of AI and the willingness to apply and use artificial intelligence!”