Mylly.AI meets the stringent security requirements of enterprise environments. The security of Mylly.AI has been validated through deployments in the business-critical systems of listed companies and large international organisations: Anora, other listed companies and large international organisations rely on Mylly.AI for business-critical AI solutions.
How security is implemented
| Area | Implementation | |
|---|---|---|
| Environment | The customer's own Azure, AWS or GCP | |
| Data processing | All data is stored and processed within the EU | |
| Data isolation | A customer-specific, fully isolated environment | |
| User management | Microsoft Entra ID integration | |
| Auditing | All AI interactions are logged | |
| Use of models | Data is not used to train models | |
Model independence and security
Mylly.AI supports several language models and providers.The best model is selected for each use case, with security considerations guiding every choice:
- EU data centres as the primary option
- Support for Azure Data Zone deployments
- Documented language model selection principles (LLM Model Selection Principles)
Security documentation
Our commitment to information security is documented in the contractual appendix, Mylly.AI Information Security Statement. The documentation is available to customers as part of contract negotiations.
Do you have questions about Mylly.AI security?
Markku Nyman
Mylly.AI Customer Value and Product development Lead
