Skip to content

Issue with Annonymyzing Data #15

Description

@RavikumarSHaligode

Hi Team,

i tried with anonymizing data, following is the piece of code, SAP AI Core offering is getting confused with ORG Name and hence getting below response

MASKED_ORG manages AI assets in a standardized and scalable manner, supports the execution and operations of AI scenarios, allows seamless integration with MASKED_ORG solutions, and its main functions include lifecycle management of AI scenarios, data-driven decision-making, and automation of tasks using machine learning capabilities

Ideally here user will expect SAP AI Core instead of MASKED_ORG as SAP AI Core not a org in itself.

can you look into this?
Can we have provision like, Allow entities=["SAP AI Core"] and it should not mask this then

data_masking = DataMasking(
providers=[
SAPDataPrivacyIntegration(
method=MaskingMethod.ANONYMIZATION, # Choose MaskingMethod.PSEUDONYMIZATION if pseudonymization is needed
entities=[
ProfileEntity.EMAIL,
ProfileEntity.PHONE,
ProfileEntity.PERSON,
ProfileEntity.ORG,
ProfileEntity.LOCATION
]
)
]
)

input_filter = AzureContentFilter(hate=0, sexual=0, self_harm=0, violence=0)
output_filter = AzureContentFilter(hate=0, sexual=0, self_harm=0, violence=0)

def Orchestration_run(query, context, model_name):
# Configure the orchestration with templating and model
config = OrchestrationConfig(
template=template,
llm=LLM(name=model_name),
input_filters=[input_filter],
output_filters=[output_filter],
data_masking=data_masking
)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions