AI System Implementation
Tested instead of improvised
We support the selection, testing, and implementation of concrete AI systems in your company. From the initial idea in the department to the documented rollout.
Non-binding · 15 minutes · Free
How to conduct an
AI system right one?
An AI system is introduced by first testing its planned deployment and then organizing it in accordance with the EU AI Act (Regulation (EU) 2024/1689). After that, roles are clarified and the necessary organizational measures are determined. Only then do pilot operations and rollout follow.
Services
From the initial idea in the department to the documented rollout:
Vera Schmidt
Managing Director, mip Consult GmbH
Non-binding · 15 minutes · without preparation
our approach
How does the implementation of an AI system work?
From the initial consultation to the released system – in four comprehensible steps.
Step 01
Free initial consultation
We offer companies a free initial consultation by phone regarding general AI-related questions and what specific advice we can provide.
Step 02
AI Quick-Check
In a short consultation session, we analyze together which AI systems are already being used in the company, which data protection and compliance risks exist, and whether there is a need for action regarding the GDPR or the EU AI Act.
Step 03
Customized offer
Based on the AI Quick Check, we will create a personalized quote for you or recommend one of our consulting packages.
Step 04
Start of consulting
After signing the consulting agreement, our experts will begin the consulting process.
Pricing
How much does the implementation of an AI system cost?
We support the implementation of a single AI system as a separate project or manage your implementation projects on a long-term basis – tailored to the scope and number of systems.
Price on request
The scope and price depend on the complexity and risk class of the system. In the free initial consultation, we will clarify your needs; after that, you will receive a personalized quote.
Lawyers from Eggert & Partner Rechtsanwälte · Service in Germany
Frequently asked questions about the introduction of AI systems
When should we have an AI system checked?
As early as possible, preferably before procurement. The later the testing is, the more expensive changes to the contract, configuration, or process will be. For systems that are already in operation, we perform the testing and document the results in the AI register.
What does the classification under the EU AI Act mean?
The classification clarifies two questions: What role does your company play, for example as a provider or operator, and what risk class does the system fall into? Both of these factors determine the specific obligations. Without this classification, it is not possible to determine what needs to be done.
Who approves an AI system?
This determines your AI governance. It is customary for a decision to be made by the management or a designated committee, prepared by the person responsible for the technical aspects and the AI officer. It is crucial that the decision is documented.
What happens when an AI system is already in use?
Then we begin with an inventory: What is being used, by whom, with what data. Subsequently, the system is retroactively organized, evaluated, and either released, subject to conditions, or replaced. This is the rule, not the exception.
Do we need a data protection impact assessment for every AI system?
No. A data protection impact assessment in accordance with Article 35 GDPR is required if the processing is likely to involve a high risk to the rights and freedoms of data subjects. We assess whether this is the case as part of the assessment and, if necessary, conduct the impact assessment.
How long will the implementation take?
It depends on complexity, risk class, and the question of how well the foundations are in place. If an AI governance structure already exists, it goes much faster than with the first project. We will provide a reliable timeframe after the AI Quick Check.
