AI Strategy & Applications

Putting artificial intelligence to practical use

AI only creates value when teams actually use it in their daily work – in their tools, processes and decisions. The focus is on training, prototypes and concrete applications, not endless PowerPoint strategies.

Services

Holistic AI consulting – from strategy to implementation.

AI Training & Practical Application

Workshops on Claude, Perplexity, local LLMs and automation tools – tailored to your teams' daily work.
Hands-on sessions: your own documents, emails and processes as examples.
Goal: Independent, responsible use instead of 'just trying it out a bit'.

Vibe Coding & AI App Development

Development of internal tools and assistants with AI-powered 'vibe coding' (e.g. v0, Lovable, Cursor) – quickly visible prototypes that can be iteratively refined.
Building workflows and apps that can run on your on-premise NVIDIA DGX setup to keep sensitive data local.
Integration into existing processes and systems so specialist teams benefit without major IT overhead.

AI Images, Video & Sound with ComfyUI and Co.

Conception and implementation of image, video and audio workflows with tools like ComfyUI and modern generative models.
Design of visuals (e.g. campaign images, illustrations, slide assets) and simple clips and audio snippets for communication, training or internal prototypes.
Building reusable ComfyUI pipelines that fit into your existing infrastructure – on your own GPU setup if needed.

Automation with n8n

Design and implementation of n8n workflows that automatically connect AI models, data sources and existing tools.
Building robust, traceable automations with logging, error handling and clear structure instead of 'Zap spaghetti'.
Deploying AI logic directly in n8n flows, e.g. for text analysis, decision rules or document automation.

Strategic Investment Advisory (Tech Perspective)

Assessment of AI trends (models, open source, agents, infrastructure) from an investor's perspective.
Technical evaluation of startups, products and platforms: maturity, differentiation, risks.
Support with due diligence and investment decisions.

On-Premise AI Infrastructure

Use of my NVIDIA DGX Spark setup for confidential analyses – no cloud dependencies, full data control.
Building an internal on-premise NVIDIA DGX Spark AI infrastructure.

Data Sovereignty with NVIDIA DGX Spark

For organizations with high data protection requirements, I operate my own on-premise AI setup. Sensitive documents from administration, politics or regulated industries are processed locally – without external cloud.

Confidential document analysis
Local LLM inference
No data leakage to external providers
Full control and compliance

🔒

Ideal for:

• Municipal and cantonal administrations
• Political organizations
• Law firms
• Financial service providers
• Healthcare

Case Studies

Possible scenarios and solution approaches

Training

Case 1: AI enablement for a specialist team

Starting point: A specialist team wants to use AI but is unsure about quality, data protection and workflow integration. Approach: Compact training program with live applications on real tasks, complemented by team 'playbooks' and example prompts. Result: AI becomes firmly embedded in research, document creation and communication; noticeable time savings and better results.

Development

Case 2: Internal AI assistant via vibe coding

Starting point: A company wants an internal assistant for documents, FAQs and texts without a major project. Approach: Development of a lean web tool with LLM API and vibe coding stack; connection to internal knowledge sources, simple permissions, iterative fine-tuning with pilot users. Result: High daily usage, relief from routine tasks and a visible starting point for further AI projects.

Automation

Case 3: Media monitoring pipeline

Starting point: An association wants to systematically capture relevant media reports without employees having to search manually every day. Approach: Building an automated pipeline with n8n – aggregation from various sources, AI-powered summarization and daily briefings by email. Result: Time savings for the team, better information landscape and a scalable system for future monitoring requirements.