Data Storage
The two types of data that may be stored on Tensorlake is documents and parsed output. Below is the default policies around data storage.There are options for Hybrid and Fully-Disconnected On Prem usage of Tensorlake. Contact us at support@tensorlake.ai
if you have to ensure your documents and parsed data never leave your servers.
- Documents: Documents can either be uploaded to Tensorlake or provided via a link.
- Uploaded documents are stored in accordance with our data storage policy below.
- Linked documents are not stored by Tensorlake.
- Parse Output: The output from a parse job includes a markdown representation, a document layout, structured data, and page classifications.
- All of the output of your parse job is stored in accordance with our data storage policy below.
Storage Policies
- Storage Location: We utilize Amazon Web Services (AWS) S3 for storing data. Data is encrypted at rest and in transit.
- Access Permissions: Access to AWS S3 storage is strictly limited to the internal document processing services. This ensures that only authorized and authenticated processes can interact with the stored data, minimizing the risk of unauthorized access.
- Data Retention: For all users, you can delete your documents and data from our servers at any time using our APIs.
- Data Usage: For all users, we never use any of your data for training purposes. We respect the privacy of our customers and ensure only they have access to the data from their requests.
Deleting Your Data
While your data is stored securely in accordance with our storage policies outlined above, we understand that you may want to remove specific documents or parse outputs at any time. If you need to request complete data deletion and/or access audit logs from Tensorlake, please contact us at support@tensorlake.ai .- Document Deletion: You can delete any uploaded document from our servers using the document ID (
doc_id). Once deleted, the document is permanently removed from our storage and cannot be recovered. The output of any parse job that referenced this document will still be accessible. - Parse Output Deletion: You can delete the output from any parse job using the parse ID (
parse_id). This removes all associated data including the markdown representation, document layout, structured data, and page classifications. Deleting the parse output will not delete the document that was parsed. - Immediate Deletion: When you request deletion, the data is immediately removed from our active systems. This ensures that you maintain full control over your data lifecycle.
- API Access: Data deletion can be performed through our API endpoints, allowing you to integrate data management into your workflows and compliance processes.
Encryption
- Encryption at Rest: All data stored in AWS S3 is encrypted at rest using industry-standard encryption algorithms. This means that even if unauthorized individuals were to gain access to the stored data, they would not be able to decipher it without the proper encryption keys.
- Encryption in Transit: We employ encryption protocols to protect data in transit. All communication between our systems and the data storage is conducted over secure channels using encryption mechanisms such as SSL/TLS. This ensures that data remains confidential and tamper-proof during transmission.
List of Authorized Subprocessors
| Company | Description | Country (where subprocessing takes place) |
|---|---|---|
| Amazon Web Services, Inc. (AWS) | Cloud Infrastructure | United States, EU |
| OpenAI, LLC | Artificial Intelligence | United States, EU |
| Anthropic PBC | Artificial Intelligence | United States, EU |
| Datadog | Error Monitoring | United States |
| PostHog, Inc. | Product Analytics | United States |
| Google Cloud | Cloud Infrastructure | United States, EU |
| Microsoft Azure | Cloud Infrastructure | United States, EU |
| Lambda Labs. | Cloud Infrastructure | United States, EU |