Unstructured extraction

Unstructured extraction extracts data from unstructured content and converts it into structured content in Reltio Context Intelligence Platform. It extracts data from files such as PDF, PPTX, DOCX, HTML, RTF, TXT, JPEG, and PNG files, and maps the extracted data to Reltio entities for review and downstream use such as golden record creation, segmentation, and enhanced context for autonomous agents. The feature also supports human in the loop reviews, so you can review and correct extracted values before the data is used for downstream processes.

Unstructured search

Unstructured search makes information contained in documents searchable by agents. It processes documents from a supported source and creates a knowledge base that an agent can query. Unstructured search divides the content into smaller passages, known as chunks, and creates numerical representations, known as embeddings. The index uses these representations and document metadata to locate relevant passages without sending every source document to the agent.

Why would I use Unstructured augmentation?

Many business processes rely on information stored in documents such as contracts, onboarding forms, invoices, and compliance records. When this information stays in documents, teams often have to review it manually and enter the data into business systems. This slows down work, increases the risk of errors, and makes it harder to use document data alongside structured master data.

Unstructured extraction in Unstructured augmentation helps solve this by extracting key information from documents and converting it into data that can be reviewed and mapped to Reltio entities. You can review and validate this data before creation of the entities and their attributes in Reltio.

Unstructured search creates searchable knowledge bases that your custom agents and autonomous agents agents can query through natural-language questions.

Unstructured augmentation provides the following benefits:

• It improves productivity by automating the and enabling natural-language search across large volumes of internal document content.
• It reduces reliance on manual document review and data entry, so teams can focus on higher-value work.
• It speeds time to value by reducing the need for extensive configuration and custom integration.
• It makes unstructured document content usable in Reltio by linking extracted entities, relationships, and attributes to master records.
• It provides richer context for AI, automated workflows, and analytics by making document-derived data available alongside structured master data.
• It makes information across documents searchable by organizing the content into knowledge bases.
• It enables agents to ask questions in natural language and receive answers grounded in the source documents.

Watch how Unstructured extraction works

The following video explains how to create a template, configure a source, and run a document pipeline in Unstructured extraction.