AI glossary for content assistants
Plain-English definitions of 13,917 AI terms for branded assistant teams.
Search glossary terms
13,917 glossary pages match your filters.
Category
Browse by letter
Glossary
13,917 terms. Open one for definitions and related concepts.
OpenAI Embedding 3 Small
OpenAI's cost-efficient embedding model that produces high-quality vectors with configurable dimensionality from 256 to 1536.
OpenAI Embedding 3 Large
OpenAI's highest-quality embedding model with configurable dimensionality up to 3072, designed for applications requiring maximum retrieval accuracy.
BGE-M3
A versatile open-source embedding model supporting multiple languages, retrieval modes (dense, sparse, and multi-vector), and input lengths up to 8192 tokens.
E5-Mistral
A high-performance embedding model built on the Mistral-7B language model, achieving state-of-the-art retrieval quality through instruction-tuned training.
Nomic Embed
An open-source, high-performance embedding model with a fully auditable training pipeline and competitive quality across retrieval benchmarks.
Arctic Embed
Snowflake's open-source embedding model family optimized for enterprise retrieval, offering multiple sizes from lightweight to high-accuracy variants.
All-MiniLM
A compact, fast sentence embedding model from the Sentence Transformers library, widely used for lightweight semantic search and similarity tasks.
ColBERTv2
An improved version of ColBERT that uses residual compression to drastically reduce the storage requirements of multi-vector retrieval while maintaining quality.
Learned Sparse Embedding
Sparse vector representations generated by neural models that learn which terms are most important, outperforming traditional keyword-based sparse methods.
Late Interaction Embedding
An embedding approach where query and document are encoded independently but compared through fine-grained token-level interaction at search time.
Angular Distance
A distance metric that measures the angle between two vectors in embedding space, related to cosine similarity but expressed as an angular measurement.
Maximum Inner Product Search
A search method that finds vectors with the highest dot product to a query vector, useful when vector magnitudes carry meaningful information.
Similarity Threshold
A configurable cutoff value that determines the minimum similarity score required for a retrieved document to be included in RAG context.
Recursive Text Splitting
A chunking strategy that recursively divides text using a hierarchy of separators, trying larger natural boundaries before falling back to smaller ones.
Markdown Chunking
A structure-aware chunking method that splits markdown documents along headings, code blocks, and other structural elements to preserve document organization.
HTML Chunking
A chunking approach that parses HTML document structure to split content along semantic boundaries defined by HTML tags and elements.
Code Chunking
A specialized chunking method for source code that splits along syntactic boundaries like functions, classes, and modules to preserve code structure.
Proposition Chunking
A chunking method that breaks text into self-contained factual propositions, each expressing a single complete claim or piece of information.
Contextual Chunking
A technique that enriches each chunk with surrounding context or document-level summaries so chunks remain meaningful when retrieved in isolation.
Chunk Metadata
Structured information attached to each chunk such as source document, page number, section heading, and creation date, used for filtering and context.
Pre-Filtering
Applying metadata-based filters before vector similarity search to narrow the candidate set, improving both relevance and search performance.
Post-Filtering
Applying metadata-based filters after vector similarity search to refine results, simpler to implement but potentially less efficient than pre-filtering.
Cohere Rerank
Cohere's neural re-ranking API that scores query-document relevance using a cross-encoder model, dramatically improving retrieval precision in RAG pipelines.
Two-Stage Retrieval
A retrieval architecture that combines fast initial candidate selection with a slower, more accurate re-ranking step to optimize both speed and quality.
Query Classification
The process of categorizing incoming queries by intent, type, or topic to route them to the most appropriate retrieval strategy or data source.
Query Routing
Directing queries to different retrieval strategies, knowledge sources, or processing pipelines based on query characteristics and classification.
Multi-hop Retrieval
Multi-hop retrieval answers complex questions by performing multiple sequential retrieval steps, where each step uses the previous result to formulate the next query.
Parent Document Retrieval
Parent document retrieval indexes small chunks for precise matching but returns their larger parent document as context, combining retrieval precision with response quality.
Contextual Compression
Contextual compression filters and compresses retrieved documents to extract only the most relevant portions before passing them to the LLM, reducing context length and improving answer quality.
Metadata Filtering
Metadata filtering narrows vector search by pre-filtering documents based on structured attributes like date, author, category, or source before semantic similarity comparison.
TF-IDF
TF-IDF (Term Frequency-Inverse Document Frequency) is a classic information retrieval algorithm that scores document relevance based on word frequency, used in RAG systems as the basis for sparse keyword search.
Embedding Models
Embedding models convert text into dense numerical vectors that capture semantic meaning, forming the foundation of semantic search and retrieval in RAG systems.
Annoy
Annoy (Approximate Nearest Neighbors Oh Yeah) is an open-source library by Spotify that builds static tree-based indexes for fast approximate nearest neighbor search in high-dimensional vector spaces.
Chunking Strategies
Chunking strategies are methods for splitting documents into segments for RAG indexing, with the choice of strategy significantly affecting retrieval precision and response quality.
AI Agent
An AI agent is an autonomous system that can perceive its environment, make decisions, and take actions to achieve goals, often using tools and integrations.
Autonomous Agent
An AI agent that operates independently with minimal human intervention, making its own decisions about which actions to take to achieve a given goal.
Semi-autonomous Agent
An AI agent that can take independent actions within defined boundaries but requires human approval for important decisions or high-risk operations.
Reactive Agent
An AI agent that responds directly to current inputs without maintaining internal state or planning ahead, acting based on immediate stimulus-response patterns.
Proactive Agent
An AI agent that anticipates needs and initiates actions without waiting for explicit requests, acting on predictions about what will be helpful.
Deliberative Agent
An AI agent that maintains an internal model of its environment and uses explicit reasoning and planning to decide on actions before executing them.
Cognitive Agent
An AI agent modeled on human cognitive processes, incorporating perception, reasoning, learning, memory, and decision-making in an integrated architecture.
Conversational Agent
An AI agent specialized in natural language dialogue, maintaining context across multiple turns and engaging in coherent, helpful conversations with users.
Task-oriented Agent
An AI agent designed to accomplish specific tasks like booking appointments, placing orders, or resolving support tickets through structured dialogue and actions.
Research Agent
An AI agent that autonomously gathers, analyzes, and synthesizes information from multiple sources to produce comprehensive research outputs.
Coding Agent
An AI agent that can write, modify, test, and debug code autonomously, often integrated with development tools and version control systems.
Planning Agent
An AI agent that creates structured plans for accomplishing complex goals, breaking them into ordered steps before executing them.
Web Agent
An AI agent that can navigate and interact with websites, reading page content, clicking buttons, filling forms, and extracting information from the web.
Browser Agent
An AI agent that controls a web browser to perform tasks, interacting with web pages through clicks, typing, scrolling, and navigation just as a human would.
Turn owned content into answers
Use InsertChat to launch a branded assistant visitors can ask directly.
7-day free trial · No card required
Try the FAQ like a visitor.
Open product, pricing, security, integration, and free-tool questions in the same chat your visitors use.
InsertChat
Interactive FAQ
Hey. Pick a question below and see how InsertChat turns FAQs into clear, source-backed answers.
Product FAQ
What is InsertChat?
InsertChat is a white-label AI assistant for your website. Train it, brand it, publish it, and learn from visitor questions.
How does InsertChat use my website content?
Connect approved pages, docs, videos, FAQs, policies, and other sources. InsertChat turns them into source-backed answers and next steps.
Can I control the assistant's tone and sources?
Yes. Choose its sources, tone, welcome message, and prompts so it stays on brand.
How does InsertChat stay accurate?
Answers use approved content and source links. Analytics show unclear or missing answers so you can improve coverage.
Can it collect leads or route support questions?
Yes. InsertChat can collect details, qualify intent, add context, and send chats to the right inbox, CRM, workflow, or person.
Can I control how the assistant behaves?
Yes. Control prompts, model choice, tool access, and the branded assistant experience so behavior stays consistent.
Which AI models can I use?
InsertChat supports multiple model providers. Choose each assistant's model for quality, speed, and cost, or use BYOK.
Can I pick different models for different workflows?
Yes. Use a faster model for common questions and a stronger model for complex reasoning. InsertChat supports that balance per conversation.
Where can I deploy an assistant?
Use a widget, embed, full-page assistant, custom domain, in-app embed, or API. Reuse one setup across surfaces.
Do I need coding skills?
No. Build and deploy AI assistants using our visual builder. The embed code is one line of JavaScript.
Can I customize the branding and UI?
Yes. Customize the assistant name, logo, colors, welcome message, suggested prompts, tone, domain, and white-label presentation.
Can I use my own domain?
Yes. Custom domains are supported, typically via enterprise options.
Does InsertChat support voice?
Yes. Voice dictation and text-to-speech let users speak instead of type.
Does InsertChat support vision?
Yes. Enable vision for assistants when images help clarify a request or context.
What tools and integrations are supported?
Zendesk, HubSpot, Shopify, WooCommerce, calendar booking, web search, Perplexity, and webhooks for your own systems.
Can I control which tools the assistant is allowed to use?
Yes. Tool access is controlled per assistant so you enable only what you need.
Can the agent hand off to a human?
Yes. Configure human handoff so the agent escalates when needed. Full conversation history is passed along.
Do you provide analytics?
Yes. Track chats, leads, feedback, top questions, unanswered questions, most-used sources, and content gaps.
Is it mobile friendly?
Yes. The widget and embeds work well on desktop and mobile with no separate experience needed.
What's the fastest path to a successful deployment?
Start with one assistant and a small set of high-value sources. Iterate using real questions from analytics.
What is the fastest way to get started?
Create an account. Connect one key source. Ask a test question, brand the assistant, then publish it on one page.