01
Manual data transfers
Someone copies data between several systems every day. I automate the flow without replacing your existing landscape unless that is needed.
Problem solving, not platform selling
AI and automation for companiesThe right solution for your process problem.
I build custom AI and automation solutions for companies. I automate recurring flows, connect existing systems and add AI where it brings a real advantage. I only choose the tech after the process is understood.
Whether AI, Microsoft 365, custom software or classic automation: what counts is not the tool, but a setup that runs reliably and fits your data, costs and workflows.
Dominik Pilichowski, Industrial Engineer (M.Sc.)
Typical starting points
02
Searching instead of working
People search for information in files, SharePoint or emails. I build setups that make existing company data accessible in a structured way.
03
Tech without a clear benefit
You want to use AI or automation, but do not yet know where it actually makes sense. I start with the process and then pick the fitting tech.
Approach
Not the strongest model. The right one.
Many AI projects start with the question of which model to use. I start with the process.
I check where the real bottleneck sits, which data is processed and what the requirements are for cost, data protection and maintainability. Only then do I decide whether local AI, a cloud model, classic automation, Microsoft 365 or custom software is the right setup.
That way you do not get a setup that depends unnecessarily on one vendor, model or tool.
For sensitive data a local setup can make sense. If demanding models or high scale matter more, cloud AI can be the better choice. The concrete process decides.
Technology-open does not mean arbitrary. n8n, Microsoft 365, cloud APIs, local AI or custom software can each be the right basis. What counts is how they fit into the existing landscape and can be operated long term.
Pilicore Agent
Architecture proof, not a product pitch
The Pilicore-Agent is my reference architecture. It shows how I build AI and automation technically: modular, extensible and independent of a single model vendor.
The architecture connects agents, workflows, company knowledge and integrations. The underlying process stays decoupled from individual LLM vendors and tools.
This is not a product to buy. It is a concrete look at how I work and the technical direction I also follow in client projects.
Case Studies
Not just concepts. Concrete implementations.
Selected projects show how different technical problems were solved in practice. From AI-supported knowledge search to small automations that take over recurring manual work completely.
AI knowledge search for SharePoint
Search grown document bases in natural language and get answers with traceable sources.
View case studyAutomated Linux security updates
Recurring system updates automated, including reboot checks, logging and error notification.
View case study
Experience
Experience from process, software and IT
I connect business-process understanding with technical delivery. My background covers business-critical applications in insurance, complex IT landscapes in logistics, international IT projects, cloud migrations and work in regulated environments.
Context
Experience with business-critical applications, complex IT landscapes and projects with high demands on documentation, traceability and operations. Typical are mid-sized firms and companies where stability and transparent processes matter.
Technologies
Experience with different AI models, cloud and local AI, Azure, GCP, Microsoft 365, APIs, scripting and custom software development. The choice follows process, data risk, cost and maintainability.
Working style
From process analysis to a productive solution. Documented, remote, and so your team understands the setup, can operate it and later develop it further.
More on my background and how I work is on the About page.
Services
How I can help
Not always the strongest model and not always a specific platform. What counts is the setup that fits process, data risk and cost.
AI
AI assistants & knowledge
Problem: information sits scattered in documents, emails or company systems. Solution: an assistant reaches your sources in a controlled way and puts relevant information where your team works.
FX
Process automation
Problem: recurring work costs time and creates unnecessary errors. Solution: automate manual flows, connect systems and leave only the cases with people that actually need a decision.
∫
System integration & design
Problem: systems do not talk to each other, or it is unclear which tech a process actually needs. Solution: connect existing systems, assess options and design an architecture that stays maintainable long term.
Transparency
Price orientation
No packages and no artificial product tiers. The basis is a day rate, aligned with usual rates for independent IT and AI consulting. That way you can roughly gauge the possible effort before you enquire.
- Day rate
- from €1,100
- Hourly rate
- from €140
What a working day covers
- Concept work, implementation or integration, depending on the project phase
- Alignment and short status updates
- Testing and clear documentation
- Handover so your team can keep running the solution
Orientation only, not a quote. After scoping, a fixed price is also possible. The first conversation is non-binding.
Describe your process problem
A short note is enough. Describe what still runs by hand, where data is copied between systems or which flow costs unnecessary time.
I usually reply within 1 to 2 business days and say clearly whether and how I can help.