AI compass for companies: five steps to your first AI project
The AI compass gives companies and SMEs a clear direction for getting started with artificial intelligence: check the prerequisites, choose the right use case, estimate costs realistically, pick the right engagement model and find vetted AI experts. Free, neutral and jargon-free.
Free · no registration needed
Your path to the first AI project
Work through the steps in order. Each leads to an in-depth page with concrete help.
1. Check prerequisites
The free AI readiness check shows in three minutes where your company stands and what to settle before a project.
2. Choose a use case
Pick the first use case by value, feasibility, data and risk — small and measurable rather than big and vague.
3. Estimate costs
Typical day rates, project costs and the factors that drive your budget.
4. Choose a model
Freelancer, AI agency or platform — which model fits project, capacity and budget.
5. Find experts
Compare vetted AI consultants and service providers, or post your AI project as a tender.
Six questions every management team should ask before starting with AI
- Which concrete problem should AI solve — and how will we measure that it is solved?
- Which data do we have for it, and may we use it for this purpose?
- Who in the company owns the initiative and has time for it?
- Do we build, buy a standard solution or work with external experts?
- Which requirements apply for data protection, hosting and the EU AI Act?
- What happens after the pilot: who runs, maintains and improves the solution?
AI glossary: the key terms explained briefly
- Generative AI (GenAI)
- AI that creates new content — text, images, code or audio — rather than only classifying data.
- Large language model (LLM)
- A large language model that understands and generates text, such as the models behind ChatGPT, Claude or Gemini.
- RAG (retrieval-augmented generation)
- A method where a language model looks up relevant passages in your own documents before answering — with sources instead of invention.
- AI agent
- An AI system that performs several steps on its own, such as reading emails, looking up data and creating entries in systems.
- Computer vision
- AI that analyses images and video, for example for quality inspection or document recognition.
- Fine-tuning
- Further training an existing model on your own examples so it handles a specific task better.
- Proof of concept / pilot
- A small, time-boxed project that shows with real data whether an AI approach reaches the goal.
- MLOps
- Operating, monitoring and improving AI models after go-live.
- EU AI Act
- The European AI regulation, which classifies AI systems by risk and attaches obligations to each class.
A note on the name
Several German chambers of commerce (IHK) publish their own overviews of events, training and funding under the name "KI-Kompass". This AI compass is an independent guide by Die KI Plattform. Regional IHK offers and the Mittelstand-Digital Zentren are a good complement, for example for training and funding.
AI compass — common questions
What is the AI compass? +
The AI compass by Die KI Plattform is a free guide for companies starting with artificial intelligence. It leads in five steps from checking prerequisites through use-case selection, costs and the engagement model to finding vetted AI experts.
Who is the AI compass for? +
For management, business departments and IT in companies without their own AI team — especially SMEs. No prior knowledge is needed; the glossary explains the key terms.
Is this the IHK's KI-Kompass? +
No. Several IHKs publish their own overviews named "KI-Kompass", for example on events and training in their region. This guide is independent and complements such offers with concrete steps through to implementation.
Where should I start with AI? +
With a clearly defined problem and an honest look at the prerequisites. The free AI readiness check shows in three minutes whether a pilot makes sense or whether use case and data should be clarified first.
What does getting started with AI cost? +
A well-scoped proof of concept often starts at €5,000–20,000. Typical day rates for AI experts range from €800 to €1,800. Costs depend mainly on data, integration and data-protection requirements.
Do I need a consultant to get started with AI? +
Not necessarily, but external experience usually pays off for choosing the first use case and delivering it. On Die KI Plattform you compare vetted AI consultants, freelancers and agencies for free.
Ready for the first step? Find vetted AI experts — for free.
Describe your project or post it as a tender and get matched with AI consultants and agencies across the DACH region. No placement fee.