Learn the language of automation.
A simple guide to the terms we use every day. Clear explanations to help you understand how the technology works.
AEO
Answer Engine Optimization: helping AI chat and search surfaces return clear, source-backed answers.
Tags: Search, MarketingAgents
Software that uses an LLM plus tools to plan and execute tasks toward a goal.
Tags: Basics, SystemsAI Alignment
The field of ensuring AI systems do what humans actually want them to do.
Tags: SafetyAI Analytics
Measuring how an AI system performs so you can improve quality, cost, and safety.
Tags: Analytics, BasicsAI Automation
Automations powered by models that can read, decide, and act with your data.
Tags: Automation, SystemsAI Chat Interface
A chat-shaped surface that lets users talk to an AI inside your product.
Tags: Assistants, UXAI Compliance
Meeting legal and policy requirements when you build and run AI features.
Tags: Compliance, RiskAI Decision Making
Using models to choose or recommend actions based on data and goals.
Tags: Product, BasicsAI Deployment
The process of serving a trained model so real users and systems can call it.
Tags: Ops, BasicsAI Engine
The core service or model that powers AI features behind the scenes.
Tags: Basics, ModelingAI Governance
Policies, processes, and oversight that guide how AI is built and used.
Tags: Governance, RiskAI Hallucination
When an AI confidently makes up facts or details that are not grounded in real data.
Tags: Risks, LLMsAI Inference
Running a trained model to get an output for a real request.
Tags: Ops, BasicsAI Integration
Connecting an AI model to your product or process so it can act on real data.
Tags: Basics, IntegrationsAI Knowledge Base
A living library of answers and sources that an AI can search to respond accurately.
Tags: Data, DocsAI Model
The learned function that maps inputs to predictions, decisions, or generated content.
Tags: Basics, ModelingAI Monitoring
Watching model quality, cost, and safety in production so you can react fast.
Tags: Ops, AnalyticsAI Optimization
Tuning prompts, models, and settings to improve quality, speed, and cost.
Tags: Ops, ProductAI Personalization
Using AI to tailor content and actions to each person’s context and history.
Tags: Product, BasicsAI Platform
A set of tools and services that let you build, ship, and manage AI features in one place.
Tags: Platforms, SystemsAI Prediction
Using a model to forecast an outcome or score before it happens.
Tags: Product, MLAI Recommendation
Models that suggest the next best item, action, or piece of content.
Tags: Product, BasicsAI Search
Search that understands meaning, not just keywords, to surface the best answer fast.
Tags: Search, RAGAI Security
Protecting models, data, and outputs from abuse, leaks, and tampering.
Tags: Security, GovernanceAI Software
Software that bakes AI into the product so users get smarter help by default.
Tags: Basics, ProductsAI System
All the pieces that make an AI feature work: models, data, tools, and guardrails together.
Tags: Basics, SystemsAI Temperature
A setting that controls how predictable or creative an AI model’s outputs are.
Tags: LLMs, SettingsAI Tool
A focused app or feature that uses AI to solve one job well.
Tags: Basics, AssistantsAI Training
Teaching a model with data and feedback so it learns patterns that generalize.
Tags: Modeling, BasicsAI Wrapper
An application that adds a user interface on top of an existing AI model.
Tags: ProductsAlgorithmic Bias
When AI creates unfair outcomes because of prejudices in its training data.
Tags: SafetyAPI
A set of rules that let one piece of software talk to another.
Tags: Basics, IntegrationsArtificial General Intelligence (AGI)
A hypothetical AI that can learn and solve any intellectual task a human can.
Tags: ConceptsAugmentation
The process of using AI to enhance human capabilities and productivity rather than replacing them.
Tags: Strategy, Human-AIAutomation
A set of steps that run on their own so work moves forward without manual clicks.
Tags: Basics, SystemsChain of Thought (CoT)
A prompting technique where the AI explains its reasoning step-by-step before giving an answer.
Tags: PromptingChatbot
A conversational interface that lets people talk to software in plain language.
Tags: Basics, AssistantsComputer Vision
Teaching software to interpret images and video so it can detect, classify, and describe what it sees.
Tags: Basics, MLContext Caching
Saving the 'memory' of a long conversation or document so you don't have to re-process it every time.
Tags: OptimizationContext window
The maximum amount of tokens an LLM can consider at once (input + output).
Tags: Basics, LLMsCopilot
An AI assistant that works alongside you to suggest improvements or automate tasks.
Tags: BasicsCost Structure
The transformation of business expenses from variable labor costs to fixed, scalable software costs.
Tags: Business, EconomicsCRM
Software that tracks contacts, deals, and customer history in one place.
Tags: Business, ToolsData Source
The place your AI or app pulls truth from: a database, CRM, wiki, or file store.
Tags: Data, BasicsDeep Learning
A subset of machine learning that uses multi-layered neural networks to learn from vast amounts of data.
Tags: MechanicsDeepfake
Synthetic media where a person's likeness is convincingly swapped or generated by AI.
Tags: SafetyFew-Shot Prompting
Giving the AI a few examples of what you want before asking it to do the task.
Tags: PromptingFine-tuning
The process of training a pre-existing AI model on specific data to make it an expert.
Tags: MechanicsFoundation Model
A large-scale AI model trained on vast data that serves as a base for many applications.
Tags: ConceptsFunction Calling
The ability of an AI to connect to external tools and execute code rather than just talking.
Tags: CapabilitiesGenerative AI
Models that create new text, images, code, audio, or video from patterns they learned.
Tags: Basics, LLMsGEO
Generative Engine Optimization: tuning content so generative search and chat pick it up accurately.
Tags: Search, MarketingGPU Cluster
Thousands of graphics cards linked together to act as one supercomputer for training AI.
Tags: HardwareGrounding
Anchoring an AI's responses in real-world facts or specific data to prevent fabrication.
Tags: ReliabilityInference
The moment the AI actually 'thinks' and produces an output after you hit enter.
Tags: MechanicsInformation Asymmetry
The competitive advantage gained by businesses that use AI to process, analyze, and act on data faster than their competitors.
Tags: Economics, StrategyLabor Substitution
The direct replacement of human tasks and roles with automated AI systems.
Tags: Strategy, EconomicsLatency (Time to First Token)
The delay between sending a request and seeing the first word of the response.
Tags: PerformanceLLM
A Large Language Model: a neural network trained to predict the next token and generate text.
Tags: Basics, LLMsLLM Marketing
Using language models to plan, create, and personalize marketing at scale.
Tags: Marketing, LLMsLoss Function
The mathematical 'scorecard' an AI uses during training to measure how wrong its guess was.
Tags: TrainingLPU (Language Processing Unit)
Thinking chips designed specifically for running LLMs at extreme speeds.
Tags: HardwareMachine Learning
Using data and math to let software learn patterns instead of hardcoding rules.
Tags: Basics, MLMixture of Experts (MoE)
A architecture that picks the best specialized 'expert' models for each part of a query.
Tags: ArchitectureMultimodal AI
AI that can understand and generate multiple types of media, like text, images, and audio.
Tags: ConceptsNarrow AI (Weak AI)
AI designed to perform a specific task very well, but unable to do anything else.
Tags: ConceptsNatural Language Processing
Techniques that let software understand, generate, and work with human language.
Tags: Basics, MLNeural Network
A computer system designed to mimic the way the human brain learns and processes information.
Tags: MechanicsParameters
The internal variables or 'settings' learned by the model during training.
Tags: ArchitecturePrompt
The input instructions and context you give an LLM.
Tags: Basics, PromptingPrompt Engineering
The skill of crafting inputs to get the best possible output from an AI model.
Tags: SkillsSentiment Analysis
AI that determines the emotional tone behind a piece of text.
Tags: ConceptsSEO
Search Engine Optimization: helping web content rank and get discovered in traditional search.
Tags: Search, MarketingSLM (Small Language Model)
A compact AI model designed for efficiency and specific tasks rather than general knowledge.
Tags: ModelsSupervised Learning
Training AI using labeled examples, like a teacher with an answer key.
Tags: MechanicsSynthetic Data
Fake data created by AI to train other AI models.
Tags: TrainingSystem Prompt
The hidden 'master instruction' given to an AI that defines its persona and rules.
Tags: PromptingTokens
The chunks of text an LLM reads and generates; cost and limits are usually token-based.
Tags: Basics, LLMsTransformer
The groundbreaking architecture introduced by Google in 2017 that made modern Generative AI possible.
Tags: HistoryTuring Test
A test to determine if a machine can exhibit behavior indistinguishable from a human.
Tags: History