From idea to production AI

Who it's for: Managing directors and decision-makers who want to identify AI potential in their own business before investing in tools or platforms.

What you'll find: Our approach from the no-strings intro call through structured process capture (Camunda/BPMN) to the prioritised use-case list with ROI estimate — including a funding eligibility check (50% BAFA in old federal states, 80% in Lüneburg region and new federal states).

Who it's for: Businesses with several AI ideas, but no clarity on which to ship first and in what order.

What you'll find: Systematic scoring of every use case by effort, value, data readiness and risk. The result is a concrete plan with quick wins, milestones, ownership and ROI per use case — not a 50-page document.

Who it's for: Teams who want an AI assistant to take over recurring tasks in customer service, sales or back office — without handing data to OpenAI, Google or Amazon.

What you'll find: Custom agents built on Claude API, Ollama with Gemma 4, n8n orchestration and MCP tools. From use-case workshop through prototype (2–3 weeks) to production deployment on your own servers, with defined KPIs.

Who it's for: Businesses with digitised but not automated workflows — invoice capture, quoting, email routing, document validation.

What you'll find: Existing systems connected with n8n and Camunda workflows. Step-by-step rollout instead of big bang, AI models for classification and generation. Typical examples from production setups: daily content workflow, OAuth token refresh, parallel worker pools for data scans, cron and webhook triggers with error handling.

Who it's for: Leadership, data protection and compliance roles in SMBs that already use AI or want to introduce it, and that have to align the EU rules (AI Act + GDPR).

What you'll find: Inventory of all AI systems, classification into AI Act risk classes, gap analysis, and a pragmatic compliance framework with templates and checklists. Self-hosted architecture as a compliance lever — data stays in-house.

Who it's for: Teams and leaders who want to use AI tools in daily work — from operations to the management board, industry-specific.

What you'll find: AI literacy per Art. 4 AI Act (mandatory since February 2026), prompt engineering, safe handling of cloud and self-hosted models. Modular library in three depth levels, hands-on in sandbox environments, train-the-trainer with AI champions per department.

Who it's for: Owners and decision-makers in mid-sized trade or SMB businesses who want to picture what an AI configuration concretely looks like — before deciding for themselves.

What you'll find: Inventory at an electrical business with ten workstations: three AI-relevant roles (management, on-site lead, back office) get different tool mixes — frontier cloud model, RAG knowledge base, self-hosted language model with n8n workflows — each justified by volume, data sensitivity and compliance. Plus architecture spine, model comparison, compliance documentation and update discipline.