Data sovereignty
We believe AI innovation should never come at the expense of your organization's privacy.
Last updated: June 2026
You own your data
We claim no ownership over the data you upload or the outputs generated by our platform.
No model training
We do not use customer data, files, prompts, or outputs to train, improve, or fine-tune our shared foundation models.
Tenant isolation
Each customer tenant is fully isolated — users from one tenant have no access to data belonging to other tenants.
Our Commitment to Data Sovereignty
Data sovereignty is a core architectural principle at saiwa.ai. In practice, this means:
You Own Your Data
We claim no ownership over the data you upload or the outputs generated by our platform.Your Data Does Not Train Our Models
We do not use customer data, files, prompts, or outputs to train, improve, or fine-tune our shared foundation models.Tenant Isolation by Design
Each customer tenant is fully isolated. Users from one tenant have no access to, or interaction with, data belonging to other tenants.
Clear Data Policy
We aim to be transparent about how data is handled within our platform.
Collection & Processing
We only process the data you explicitly provide in order to deliver our AI services. We do not collect hidden or unrelated data.Encryption in Transit
All data transmitted to and from our platform is encrypted using HTTPS (TLS).File Encryption at Rest
Files stored in our infrastructure are encrypted at rest using Amazon S3 server-side encryption (AES-256).Data Deletion & Retention
- When you delete files from the platform, they are no longer accessible through the application.
- Deleted files may remain in backup or recovery systems for up to 30 days before permanent removal.
Secure & Isolated Architecture
We design our infrastructure to minimize cross-customer risk and ensure logical separation.
Single-Tenant Logical Isolation
Each tenant operates in a logically isolated environment. There is no cross-tenant data access.Operational Security
Access to production systems is restricted and controlled according to internal operational policies.