ConsultingServices.aiAI Consulting for SMEsFAQ
Frequently Asked Questions — answered honestly.
Before you inquire, you want to know what to expect. Here are the questions that come up most often — about costs, process, prior knowledge, and guarantees.
Who is the offer intended for?
For managing directors, operations managers, and sales managers of SMEs with 10–200 employees who want to use AI in a comprehensible way without a large project.
What does AI consulting cost?
The entry with the AI initial consultation is free. Packages start at €2,900 for a focused use case with a pilot phase. Each package clearly states in advance what you will receive.
Do we need prior knowledge of AI?
No. Each package is explained in such a way that decision-makers without technical prior knowledge can immediately see what is included, what it costs, and what benefits it brings.
How quickly will we see results?
Small automations often show results within 1–2 weeks. A clean pilot (starter package) delivers a testable result in 2–3 weeks.
What happens if the pilot does not work?
If the pilot does not deliver measurable results, I will not issue an invoice for the accompanying phase. The risk is on me, not on you.
Can the solutions be expanded later?
Yes. The structure is modular. New use cases, workflows, or assistants can be seamlessly added — documented in the technical handover report.
How does the AI initial consultation work?
You fill out the short form, and I will get back to you within 48 hours. In the 45-minute video call, we analyze your bottlenecks. Afterwards, you will receive a written recommendation via email — free and non-binding.
Do you work with specific tools or providers?
I am not tied to any provider. The choice of tools depends on your use case, your existing systems, and your budget — not on partnerships or commissions.
Do we need to rebuild our existing IT infrastructure for AI integration?
No. We use modern integration technologies (such as workflow platforms and middleware) that serve as a bridge between your existing systems (ERP, CRM, email) and the AI. This means: No expensive IT large projects, no vendor lock-in, and seamless integration into your current workflow.
What distinguishes you from an agency?
You have a single point of contact, no handovers. No project manager, no junior consultant — I accompany you personally from analysis to go-live.
What is the difference between ChatGPT (LLM), RAG, and an AI agent?
A regular **LLM** (like ChatGPT) is brilliant as a text generator but is based solely on its general training. **RAG** expands this concept by having the AI specifically search your company documents — perfect for secure internal knowledge without fabricated facts. An **AI agent** goes a step further: It plans tasks, uses tools (like web search or CRM), and processes tasks in multiple stages and largely autonomously.
Does the AI sometimes invent answers (hallucinations)? How do we prevent that?
This is a valid concern with free ChatGPT. We solve this with a “RAG architecture” (Retrieval-Augmented Generation) in combination with clear “guardrails.” The bot is not allowed to freely invent answers but must formulate them precisely based on the documents you provide (FAQs, manuals). If it finds nothing, it honestly admits: “I don’t know” or refers to a human.
What about data protection? Do Microsoft or OpenAI learn from our company data?
No, we technically and contractually exclude that. We use only enterprise interfaces (like Azure OpenAI Service or GDPR-compliant LLM provision in the EU) for productive corporate use. These providers guarantee: Your inputs and documents are *not* used for training the models. Your data remains yours.
Shouldn’t we just use Microsoft Copilot or ChatGPT Enterprise directly?
Hyperscaler tools (like Copilot) are excellent “assistant tools” for individual employees at their PCs. However, they reach limits when you need a tailored customer service chatbot that writes in *your* ticket system, queries *your* specific APIs, or requires an exact “total cost of ownership” (TCO) calculation for external interactions. Specialized RAG solutions or dialog managers are often more efficient and cost-effective to scale than pure pro-user licenses.
Do you offer ongoing support after the project?
Yes. Each package includes an accompanying phase with weekly status calls and email support. Additionally, you can book monthly support.
All packages with specific deliverables and prices.
Use CasesSix typical scenarios with problem, solution, and result.
About MeA single point of contact — who is behind it.
Next Step