Business model and AI platformFrom a requirements list
From a requirements list
to a business model
The client wanted an internal tool to run their ERP instances. What emerged was a white-label platform with a licence model, including an AI agent with zero-data-retention architecture.
Vertriebsdienstleister, Frankfurt
6 weeks
to go-live
Zero
data retention at the model provider
The case
Where we started.
Situation
On the table was a requirements list for an internal tool: set up and run ERP instances for clients with less manual effort. A standard delivery project, both sides initially assumed.
Brief
We suggested stepping back before building: if this tool works, it is no longer an internal tool but a product. What is missing for it to be sellable?
Approach
What we did.
White-label platform instead of an in-house tool
The architecture was designed as multi-tenant from the start, so the client can offer the platform under their own brand to their own clients.
AI platform agent with a human approval step
Natural language in, an execution plan out, then human approval, only then execution. That approval step is what makes it safe to let an agent near production systems.
GDPR architecture, not a GDPR chapter
Zero data retention at the model provider, masking of personal data before model access, a complete DPA chain, restore tests under Art. 32 GDPR. Data protection as an architectural decision, not a paragraph added at the end.
Licence model with contribution margin calculation
SaaS licence model including a per-client contribution margin calculation. That meant it was clear before the first contract at what point the platform would pay for itself.
Results
What we delivered.
- An internal tool became a sellable platform product
- Six weeks from requirements list to go-live
- AI agent with human approval before every action
- Zero-data-retention architecture with a complete DPA chain
- Per-client contribution margin calculation as the basis for pricing
„We came with a requirements list and left with a business model. The fact that both came from the same people who then built it was the real difference.“
Technology used
- Odoo
- Python
- PostgreSQL
- Docker
- LLM-Integration