Services

AI Consulting for Small and Medium-Sized Businesses

AI consulting for SMEs means finding out which tasks in your business artificial intelligence can reliably take over today — and then implementing exactly those solutions in a GDPR-compliant way. Yurtbay.dev provides AI consulting for small and medium-sized businesses, based in Recklinghausen, Germany, working on site across the Ruhr area and remotely: from potential analysis and tool selection to implementation and training. Consulting and implementation come from one person — the same person who builds the result.

How do we roll out Microsoft 365 Copilot properly?

Copilot lives inside Word, Outlook, Excel and Teams — introduced properly, it saves time every day on emails, meeting notes, proposals and reports. For that to work, permissions, data storage and licensing have to be right first: Copilot sees everything the individual user is allowed to see, including accidentally shared folders. I guide the rollout from a readiness check through a small pilot group to team training, so the licence cost actually turns into time saved.

How do we get a GDPR-compliant chatbot on our website?

A website chatbot answers routine questions about services, appointments and processes around the clock — noticeably relieving your phone line and inbox, especially when nobody should be paid to answer the same questions over and over. It becomes GDPR-compliant through EU hosting, a data processing agreement (DPA) with the AI provider and a clean privacy policy. I build chatbots that only answer from your own, vetted content — instead of making things up.

How do we automate documents, invoices and receipts?

Manually copying incoming invoices, delivery notes and forms into accounting or ERP systems costs hours every week — and produces typos. AI-powered document processing extracts amounts, line items and senders automatically and hands them to your systems in a structured way. That noticeably cuts routine work — and fits well with e-invoicing mandates, which enforce structured invoice data anyway.

How do we make our company knowledge searchable with AI (RAG)?

Proposals, manuals, project folders, old emails: the knowledge exists in your business, but nobody can find it quickly. With Retrieval-Augmented Generation (RAG), your team asks questions in plain language — “How did we solve this for client X in 2023?” — and gets answers with source references from your own documents. If required, this runs entirely on your own hardware, so no data ever leaves the building.

The entry point: AI Potential Workshop — 1 day, result: your actionable AI roadmap

You don't have to start with a big project. The AI Potential Workshop is a clearly scoped one-day format — on site at your business or remote. You get a fixed quote up front instead of hourly-rate surprises, and at the end a result you can work with immediately.

What you take away from the workshop:

Process analysis: We walk through your workflows together and identify where AI saves measurable time — and where it doesn't.
Tool recommendations: concrete tools including a GDPR assessment and a realistic cost estimate, matched to your existing IT.
Quick-win prioritisation: what pays off first, what can wait, and what you can skip entirely.
AI roadmap: a written, prioritised implementation plan — usable with me or on your own.

Public funding programmes for consulting and digitalisation projects may be an option — whether and which ones fit your business is something we check together.

GDPR-compliant AI: EU hosting, Azure OpenAI or fully on-premise

For SMEs, the choice of AI solution is almost always decided by data protection. That's why the GDPR assessment is part of my work from day one — not an afterthought. The three usual routes:

EU hosting: AI services with servers and a contracting party in the EU — when cloud is fine in principle, but transfers to the US are not.
Azure OpenAI: GPT models via Microsoft Azure with processing in the EU — according to Microsoft, your inputs are not used to train the models. The pragmatic route for businesses already on Microsoft 365.
On-premise / self-hosting: open models on your own hardware for data that must never leave the building — somewhat more effort, full control.

Every solution comes with a data processing agreement (DPA) with the provider, clear internal rules about which data may be entered, and the matching addition to your privacy policy.

How we work together — in 4 steps

Intro call (no obligation): about 30 minutes. We clarify your starting point and goals — you'll know right away whether and where AI makes sense in your business.
AI Potential Workshop: one day, together with the people who know your workflows. Result: your prioritised AI roadmap.
Implementation: I build the prioritised solutions — each as a fixed quote with a clearly defined scope.
Operations & training: team onboarding, documentation and, if you like, ongoing support so the solution sticks in day-to-day work.

No slideware AI: consulting that implements

No slideware AI: I implement what I recommend — no junior, no account manager in between. From the intro call to operations you talk to the same person who builds your solution. That also means I only recommend what I can technically stand behind — and I'll advise against AI wherever it would only add cost instead of value.

Frequently asked questions about AI consulting

What does AI consulting cost for a small business?

You get a fixed quote before every step instead of hourly-rate surprises. As market orientation: one-day workshops from small providers typically range between €1,500 and €2,500. What your specific project costs is something we clarify in a no-obligation intro call.

Can ChatGPT and similar tools even be used in a GDPR-compliant way?

Yes — if the framework is right: a data processing agreement (DPA) with the provider, data processing in the EU where possible, clear internal rules about which data may be entered, and a note in your privacy policy. For sensitive data there are alternatives such as Azure OpenAI with EU processing or self-hosted models.

Is AI already worthwhile for a business with 5 or 10 employees?

Yes, often especially there: in small teams, routine work — writing proposals, transferring receipts, searching for information — eats proportionally the most time. What matters is not company size but whether recurring tasks exist. That's exactly what the process analysis in the workshop checks.

Do our data have to go to the cloud?

No. Many solutions — such as searchable company knowledge via RAG — run entirely on your own hardware (on-premise). Whether cloud with EU processing or your own server is the right choice depends on your data, your IT and your budget; that's precisely what the AI Potential Workshop clarifies.

How quickly will we see first results?

The AI Potential Workshop takes one day and immediately delivers a prioritised roadmap. First quick wins — such as text templates, email drafting or a Copilot pilot group — can often be implemented within a few weeks. Larger projects such as document automation or RAG are planned realistically in the roadmap.

Is public funding available for AI consulting?

Funding programmes at federal and state level in Germany may be an option for consulting and digitalisation projects — but conditions change regularly and depend on the individual case. Whether and which programmes fit your business is something we check together in the intro call.