Questions & answers
Answers to common questions about MCP integrations, project planning and getting started.
Before the first step.
What does MCP add to an existing API?
An API exposes a system’s functions. An MCP server can make selected functions available as described tools and information for a compatible AI client. The existing API can remain the technical foundation.
Do we need our own AI application first?
Not necessarily. We first assess which assistant your team wants to use and what MCP functionality it supports. That determines whether an integration is sufficient or a custom interface makes sense.
What is meant by an MCP app?
In addition to data and tools, supporting clients can display interactive interfaces. Whether and how these are used depends on the target platform and workflow.
How long does implementation take and what does it cost?
This depends on existing interfaces, access permissions and the required functionality. After reviewing the technical situation, we can define the initial project phase and estimate the effort.
Will our data stay in Austria or the EU?
Hosting location alone cannot answer this. All participating services and data flows matter. We clarify your requirements and review the technical architecture with your responsible team.
Which AI platform is right for us?
We choose based on your requirements and existing tools. We assess supported functions, authentication, data processing and interfaces for the specific platform.
Can we start with a small pilot?
Yes. A clearly bounded workflow with defined data and users is a practical starting point. You can then decide which further functions are needed in operation.
Let’s discuss your project.
Tell us what you want to solve. Together we will identify the right starting point.