Frequently Asked Questions

Lume is an AI-powered data mapping platform that automates the process of transforming data between different schemas. It helps you:

• Normalize data from multiple sources
• Ingest client data automatically
• Create and maintain data pipelines
• Map data in minutes rather than days or weeks

1. Contact Lume to set up your account and get API keys
2. Install the SDK for your preferred language (Python or TypeScript)
3. Define your target schema (or upload a sample CSV)
4. Start mapping your data

Check out our Quickstart Guide for detailed steps.

Yes, Lume is SOC 2 Type 1 and Type 2 compliant. We implement comprehensive security measures including:

• Data encryption at rest and in transit
• Multi-factor authentication
• Least-privilege access control
• Regular penetration testing
• Vulnerability scanning
• Secure data retention and disposal procedures

Lume follows strict data privacy principles:

• Customer data is purged when service is terminated
• Formal data retention procedures are in place
• All data transmission is encrypted
• Access to data is strictly controlled
• Regular security training for all employees

Lume currently provides SDKs for:

• TypeScript/JavaScript
• Python

Both SDKs are currently in beta. Contact support for the latest versions and features.

Lume currently supports:

• JSON
• CSV
• Excel

Contact us if you need support for additional formats.

Target schemas conform to the JSON Schema protocol. This allows you to specify:

• Data types
• Field requirements
• Validation rules
• Format specifications
• Nested structures

Generation time varies from seconds to several minutes, depending on schema complexity:

• Simple schemas: A few seconds
• Complex schemas: Up to 20 minutes

Note: This is only for initial generation. Subsequent runs using existing pipelines execute immediately.

1. Lume infers the schema from the data sample.
2. Lume uses the values of source data to create a more holistic semantic understanding of your source key definition and your source data as a whole. This helps increase mapping accuracy significantly.

Lume recreates the human process of mapping data. In this step, it is common that the source keys are not semantically meaningful. Namely, key names not be related to what the data is actually storing. Thus, it is common to analyze the data to understand what the property truly is storing. Lume goes through this understanding process via its AI.

Lume provides detailed validation error information including:

• Error type and location
• Error statistics
• Sample problematic values
• Schema path information

Review the Validation Errors guide for detailed troubleshooting steps.

To improve mapping accuracy:

1. Provide clear, detailed descriptions in your target schema
2. Use full phrases rather than single words
3. Include business context and edge cases
4. Provide larger data samples when possible
5. Use the schema validation features to enforce data quality