AI & Document Intelligence Glossary

Restricting document and information access based on user permissions.

AI systems that can autonomously plan, reason, and execute multi-step tasks.

References that trace AI-generated answers back to source documents.

When AI generates confident but factually incorrect information.

Splitting documents into smaller segments for processing and retrieval.

The maximum amount of text an AI model can process in a single request.

Automatically pulling structured information from unstructured documents.

Policies and processes for managing data quality, security, and compliance.

AI technologies for understanding, processing, and extracting information from documents.

Automatically categorising documents by type, topic, or purpose.

Converting documents into structured, machine-readable data.

A numerical representation of text that captures its semantic meaning.

AI systems designed for business use with appropriate security, governance, and integration capabilities.

Search technology designed to find information across an organisation's documents and systems.

Adapting a pre-trained AI model to perform better on specific tasks or domains.

Connecting AI responses to verified source documents to ensure accuracy.

Combining keyword and semantic search for more comprehensive results.

Running a trained AI model to generate predictions or outputs.

Finding relevant documents or passages from a large collection based on a query.

AI-powered automation that extracts, classifies, and processes data from documents.

A structured representation of entities and their relationships.

Systems and practices for capturing, organising, and sharing organisational knowledge.

AI models trained on vast text data to understand and generate human language.

The time delay between a request and response in AI systems.

Data that describes and provides information about other data.

AI that can process and understand multiple types of data like text, images, and audio.

Identifying and classifying named entities like people, organisations, and dates in text.

AI techniques for understanding, interpreting, and generating human language.

Technology that converts images of text into machine-readable text.

Running software on an organisation's own infrastructure rather than in the cloud.

Extracting text, structure, and data from PDF documents.

Cloud infrastructure dedicated to a single organisation for enhanced security and control.

Crafting effective instructions to get desired outputs from AI models.

Interpreting user search queries to determine intent and improve results.

Metrics measuring how completely and accurately a system retrieves relevant results.

Reordering search results using a more sophisticated model to improve relevance.

A technique that grounds AI responses in retrieved documents for accurate, cited answers.

Search that understands the meaning of queries rather than just matching keywords.

Data organised in a defined format like databases, spreadsheets, or JSON.

Identifying and extracting tabular data from documents while preserving structure.

The basic unit of text that AI models process, roughly equivalent to a word or word piece.

The neural network architecture that powers modern language models.

Data without a predefined format, like documents, emails, and images.

A database optimised for storing and querying high-dimensional vector data.

AI's ability to perform tasks it wasn't explicitly trained on.