AIComplianceCheckNodeEditor

Overview

The AIComplianceCheckNodeEditor is a custom editor component for managing compliance checks in the Vantage analytics and data platform. This node allows users to define a series of compliance policies, which the integrated AI evaluates against dataset rows, categorizing each as either a "PASS" or "FAIL". This tool is critical for organizations aiming to maintain adherence to legal standards, industry regulations, or internal policies.

Purpose

The primary purpose of the AIComplianceCheckNodeEditor is to provide a user interface for creating and managing compliance policies that are utilized by an AI model during data evaluation processes. This node helps organizations ensure their data practices align with compliance requirements.

Settings

Policies

1. Name: Policies
   • Input Type: Array of Objects
   • Description: This setting holds an array of compliance policies. Each policy consists of a label and a description. Adding or modifying policies directly impacts the AI's evaluations as these dictate what standards must be met for data to pass compliance checks.
   • Default Value: An empty array [].

Prompt Template

2. Name: promptTemplate
   • Input Type: String
   • Description: This text area allows users to provide a custom template that guides the AI on how to focus its evaluation. The template can incorporate placeholders for dynamic content like {{data}}, {{columns}}, and {{policies}}, which will be replaced with actual values at runtime. This influences how the AI interprets the input data based on the established policies.
   • Default Value: An empty string ''.

Batch Size

3. Name: batchSize
   • Input Type: Numeric
   • Description: This setting determines the number of rows processed at one time by the AI during compliance checks. Adjusting the batch size affects the performance and responsiveness of the AI, as larger batches may improve efficiency but could lead to increased memory usage.
   • Default Value: 10.

Focus Column

4. Name: focusColumn
   • Input Type: Dropdown (String)
   • Description: This setting allows users to select a specific column from the dataset to focus the AI's compliance checks. This can lead to faster and more precise evaluations, especially in datasets with many columns. Choosing "All columns" makes the AI evaluate the entire row.
   • Default Value: An empty string '' (indicating all columns).

How it Works

The AIComplianceCheckNodeEditor leverages the defined compliance policies to assess each row of upstream data as either compliant or non-compliant. Users can dynamically manage the policies, prompting the AI to evaluate dataset columns based on these guidelines. The functionality includes:

• Adding, updating, removing, and reordering compliance policies.
• Inserting column placeholders directly into the prompt template for customized evaluations.
• Selecting a specific column for focused compliance checks, which can enhance the processing efficiency of the AI.

Expected Data

The AIComplianceCheckNodeEditor expects data in a structured format, primarily focusing on a dataset that contains various columns. Critical data expected include:

• An array of policies which define what constitutes compliance.
• A promptTemplate string that may incorporate dynamic placeholders.
• A list of relevant upstreamColumns drawn from the dataset to facilitate policy application.

AI Integrations

The integration allows for seamless evaluations by utilizing underlying AI technologies that evaluate each row against the compliance policies. Depending on the use case, the AI may employ natural language understanding and parsing capabilities to ensure thorough compliance checks.

Billing Impacts

Utilizing the AIComplianceCheckNodeEditor may have billing implications depending on the volume of data processed, frequency of compliance checks, and the complexity of policies defined. Higher batch sizes and frequent evaluations could potentially lead to increased processing costs. It's crucial for users to monitor usage patterns and adjust settings as needed to align costs with organizational budget constraints.

Use Cases & Examples

Use Cases

1. Regulatory Compliance in Financial Services: Organizations within the financial sector can use this component to ensure that all transactions adhere to applicable legislation and internal compliance requirements.

2. Data Privacy Maintenance: Companies handling sensitive information can set policies aimed at protecting private data, ensuring that individual consumer rights are upheld in processing activities.

3. Quality Assurance in Manufacturing: Manufacturing firms can define policies concerning raw material criteria and product specifications to ensure that every item meets industry standards.

Example Configuration

Use Case: Regulatory Compliance in Financial Services

An organization aims to ensure that its transaction records comply with anti-money laundering regulations. To set this up using the AIComplianceCheckNodeEditor, the following configuration might be employed:

Policies Configuration:

[
    {
        "label": "Customer Verification",
        "description": "All customers must be verified before transactions."
    },
    {
        "label": "Transaction Limits",
        "description": "Transactions must not exceed the allowed limit of $10,000."
    }
]

Prompt Template:

Batch Size: 50 (to efficiently handle evaluations)

Focus Column: transactionAmount (if analyzing only the amount to determine compliance with limits)

This configuration will guide the AI to focus on transaction data, leveraging defined policies to check for compliance, thus facilitating regulatory adherence.