AskTable Data Privacy Protection
Data Types
In AskTable, the main data types we process and store include:
| Image | Name | Description |
|---|---|---|
 | Chat History | Records the dialogue content between users and the AI system. These data include the questions asked by users, the answers provided by the AI, and other details of the interaction. |
 | Metadata (Table Name/Field Names) | Describes information related to data tables and fields, such as the structure of the table, field names, and their definitions. This helps in understanding the organization and relationships of the data. |
 | Agent Chain of Thought Logs | Logs the thinking and reasoning steps performed by the AI during analysis. These logs help optimize the intelligent analysis and decision-making processes of the model, ensuring that the AskTable AI can provide more accurate recommendations and answers to users. |
 | Corporate Data | Confidential data internal to the enterprise, stored in the local database of the enterprise. |
Data Storage Locations
In AskTable, there are two options for storing data:
-
Cloud:
- Data is stored on our cloud platform. Cloud storage makes data easily accessible, manageable, and backed up, while providing efficient computing power and scalability.
- We use top-tier cloud computing providers to ensure the security and reliability of the data.
-
Local:
- This refers to data being stored on the customer's local servers or private cloud. Data processing and storage are kept within the customer-controlled environment, ensuring maximum privacy protection and data security.
Data Privacy Protection Measures
Based on the four deployment modes, AskTable offers different levels of data privacy protection measures. Below is a detailed explanation for each mode:
1. SaaS
- The private data of the enterprise is still stored locally and not uploaded to the cloud. However, chat history, metadata, and agent chain of thought logs are saved on cloud servers, with the AI large model only performing inference without saving data.
- Privacy Protection Measures:
- Data Encryption: All transmitted and stored data is encrypted to ensure data security.
- Strict Access Control: Through authentication and authorization mechanisms, only authorized personnel can access sensitive data.
2. Single Machine Free Trial Version
- Corporate confidential data, chat history, metadata, etc., are all stored and processed locally. However, agent chain of thought logs generated during AskTable execution are saved on the cloud for continuous product improvement, and the cloud AI model is also called for inference (without saving data).
3. On-Premises Deployment A: Online Model
- Ensures that all data (including corporate data, chat history, metadata, agent chain of thought logs, etc.) are fully stored and processed locally, and only the cloud AI model is called for inference (without saving data).
4. On-Premises Deployment B: Local Model
- Provides the strongest data privacy protection, with all data fully stored, processed, and inferred locally, suitable for enterprises with extremely high requirements for data security.
Summary
Through multiple deployment modes, AskTable provides flexible solutions to help businesses choose the most suitable data privacy protection plan based on their own needs.
Regardless of which deployment mode you choose, AskTable is committed to ensuring your enterprise data is fully protected through strict security measures and privacy protection strategies.