migration-center Classify uses AI to identify document types and extract structured metadata from unstructured content. Build reusable analyzers to turn large document collections into reliable data for migration, compliance, search, and downstream business processes.
Prepare, enrich, classify, and validate your documents with AI to support migration and downstream business processes.
Prepare content for migration
Classify legacy documents and extract missing metadata to prepare content for a structured migration into a modern ECM, archive, or cloud repository.
Improve metadata quality
Improve incomplete or inconsistent metadata by extracting reliable values directly from document content, creating more complete and accurate data.
Route documents automatically
Assign document types automatically so documents can follow the right migration path, review process, retention logic, or business workflow.
Validate AI before production
Evaluate analyzer performance against ground-truth samples and validate classification and extraction quality before applying AI to production data sets.
Unlock scanned documents
Make scanned and image-only documents accessible to AI by using OCR to extract the text required for reliable classification and metadata extraction.
Export results for downstream use
Review classification and extraction results per document and export structured CSV or JSON data for processing in downstream systems and workflows.
migration-center Classify provides a structured workflow for building, validating, and deploying AI analyzers, guiding you from initial configuration and training through quality evaluation to controlled production processing at scale.
1. Configure & build
Configure your AI services and document sources, create an analyzer for classification, metadata extraction, or both, and define the document classes and metadata fields relevant to your use case.
2. Train & validate
Add representative documents and ground-truth data to train and evaluate your analyzer. Compare AI-generated classifications and extracted metadata with expected results to ensure the required quality before production.
3. Release & migrate
Release validated analyzers for production, process document collections at scale, review classification and extraction results, and utilize structured data for migration and other use cases.
Organizations across industries rely on migration-center for their content migration projects.
Bring structure and reliable metadata to your content before migration, move it securely to the right target, and validate the outcome afterwards. Use each edition independently or combine them to support your content from preparation through migration to final validation.
Classify documents and extract the metadata needed to prepare unstructured content for migration and other business processes.
Migrate content and metadata between content platforms while applying the required mappings and transformations.
Validate migrated content against the source and create documented evidence that migration results meet your requirements.
Bogdan Petrovan – Managing Consultant
“For me, the real value of Classify is that it makes AI usable in a controlled way. Teams can define what information they need, evaluate results against real documents, and refine the setup before processing content at scale. With support for both local and cloud-based LLMs, organizations can meet their data privacy requirements, incl. keeping sensitive data entirely on-premises. This gives teams confidence in how their data is handled and in the quality of the information they work with.”
Florian Piaszyk-Hensen – Director Products
“With migration-center Classify, built on modern AI technologies, we can intelligently identify document types and enrich content with the right metadata, access rights, and confidentiality levels. This creates a clean, reliable, secure, and governed trustful digital foundation for modern ECM systems and future AI-supported use cases – turning content migration into a strategic step toward secure, governed, and AI-ready environments.”
Get quick answers to common questions about migration-center Classify, from AI integration and document processing to quality control, production use, and how Classify fits into your existing content processes.
1. What is migration-center Classify?
migration-center Classify uses AI to identify document types and extract structured metadata from unstructured content. It helps organizations understand, enrich, and prepare large document collections for migration and other business processes.
2. What types of processing does Classify support?
Classify supports document classification, metadata extraction, or a combination of both. An analyzer can therefore identify what type of document it is processing and extract the metadata relevant to that document class.
3. Can Classify process scanned or image-only documents?
Yes. OCR engines can be configured to make text from scanned or image-only documents available for AI-based classification and metadata extraction.
4. Which AI models can I use with Classify?
Classify works with configurable LLM endpoints rather than limiting you to a single AI provider. This allows organizations to connect suitable local or cloud-based AI services according to their infrastructure and requirements.
5. How do I define what Classify should identify or extract?
You define the document classes and metadata fields relevant to your use case and describe how they should be recognized or extracted. These definitions become part of reusable analyzers that can later be applied to production document collections.
6. How can I evaluate AI results before using Classify in production?
Representative documents can be annotated with expected classes and metadata values to create ground-truth data. Classify can then compare AI-generated results against these expected values, allowing teams to evaluate quality before releasing an analyzer for production use.
7. Does Classify require AI models to be trained from scratch?
No. Classify uses configured LLM services and combines them with the document classes, metadata definitions, and instructions defined in an analyzer. Training sets and ground-truth data are used to evaluate the analyzer and its results before production use.
8. Can Classify be used independently of migration-center Migrate?
Yes. Classify can prepare and structure document data for a variety of use cases and is not limited to migration projects. When used together with migration-center Migrate, it can classify and enrich content before it is moved to a new target system.
9. What happens to the results after documents have been processed?
Classification and extraction results can be reviewed at document level before they are exported as structured output. This makes the resulting data available for migration, further processing, other systems, or business workflows.
10. How does Classify fit into the migration-center product family?
Classify complements migration-center Migrate and Verify across the content migration lifecycle: Classify can prepare and enrich content before migration, Migrate moves it to the target system, and Verify can validate migrated content afterwards. The editions address different stages and can be used according to the requirements of each project.
11. Can I use migration-center Classify with scanned documents?
Yes. migration-center Classify can work with OCR technology to make scanned or image-based documents available for AI-powered classification and metadata extraction. This allows legacy content to be processed even when no machine-readable text is available initially.
12. How can I check the quality of the AI results before using them at scale?
migration-center Classify lets you evaluate analyzers against documents with predefined ground-truth information. You can compare the AI-generated results with the expected document classes and metadata values and assess the quality before releasing an analyzer for production use.
13. Can different document types require different metadata?
Yes. With a combined classification and metadata extraction analyzer, migration-center Classify can first determine the document type and then extract the metadata relevant to that specific class. This is particularly useful for heterogeneous document collections where, for example, contracts, invoices, and reports require different information.
14. What happens with the classification and metadata after processing?
The results can be reviewed for individual documents and exported as structured data. When migration-center Classify is used together with migration-center, the classification results and extracted metadata are returned directly to the migration workflow, where they can be used for mapping, validation, correction, and ultimately the import into the target system.
Start a trial, book a demo, contact our team, or explore licensing options.
What is your target platform?
SAP SuccessFactors | d.velop documents | Microsoft SharePoint | OpenText Documentum CM | Hyland Alfresco | Box | Generis CARA | Veeva Vault | Microsoft OneDrive | File system | OpenText Content Management xECM | Sparta TrackWise Digital | OpenText Information Archive | Microsoft Azure Blob Storage | Hyland OnBase
