AiPy offers the following data security solutions:
local processing mode
- Fully offline operation: after configuring a large local model such as Ollama, all data processing is done on the user's device
- Network isolation: through
--localParameters force disabling of Internet API calls - Temporary file encryption: the generated intermediate data files are stored using AES256 encryption
privilege control
- File access whitelisting: allows you to set up access to only the data files in a specific directory
- API call auditing: all third-party API requests are logged for review
Privacy protection mechanism
- Data desensitization: automatically identify sensitive fields such as cell phone number/ID number for mask processing
- Memory cleanup: automatically clears temporary data from memory after task completion
- Configuration separation: sensitive information such as API keys are stored independently in encrypted configuration files
It is recommended that users in finance, healthcare and other industries: ① prioritize the use of local models ② use the local model before processingai('脱敏数据中的身份证号')Pretreatment ③Periodic cleaning~/.aipy_cacheCache Catalog.
This answer comes from the articleAiPy: automating the task of running Python code for data analysisThe
May not be reproduced without permission:AI productivity tools " How do I ensure data security when using AiPy to process sensitive data?
