The AI Dictionary
Every AI and automation term we use on this site, explained the way we'd explain it to a business owner over coffee. Click any highlighted term anywhere on the site to jump straight to its definition.
A
A/B testing
Comparing two versions of a page, email, or ad to see which performs better. The foundation of data-driven marketing.
Agentic AI
AI systems built around agents that plan and execute work, rather than one-shot chatbots.
AI agent
Software that doesn't just answer questions but takes actions: calling tools, checking systems, and completing multi-step tasks toward a goal.
AI receptionist
An AI that answers business phone calls: greeting callers, answering questions, booking appointments, and routing or escalating to humans.
API
Application programming interface — a documented way for software systems to talk to each other. The foundation of most automation.
Artificial intelligence (AI)
Software that performs tasks which normally require human judgment: understanding language, recognizing patterns, making predictions, or making decisions.
B
Backlinks
Links from other websites to yours. A major signal of authority to search engines.
Business intelligence (BI)
Tools and processes that turn raw business data into actionable insights through reports, dashboards, and analysis.
C
Call routing
Directing callers to the right person, department, or AI flow based on what they need.
Chatbot
A conversational interface that answers questions. Unlike an AI agent, a basic chatbot can't take actions in your systems.
Churn
Customers cancelling or not returning. Automation commonly targets the follow-up gaps that cause it.
Citations
Mentions of your business name, address, and phone number across directories. Consistency builds local ranking trust.
Computer vision
AI that interprets images and video: recognizing objects, reading text in photos, inspecting products.
Context window
How much text a model can consider at once. Larger windows let it work with longer documents and conversations.
Conversational AI
AI systems designed for natural back-and-forth dialogue, powering chatbots, voice agents, and virtual assistants.
Conversion rate
The percentage of visitors, callers, or leads who take the desired action.
Cost per lead
Total marketing spend divided by leads generated. The efficiency metric for lead generation.
CRM
Customer relationship management — software that tracks leads, customers, and every interaction with them. The memory of your sales process.
D
Dashboard
A visual display of key metrics and KPIs, updated in real time, giving a quick overview of business performance.
Data pipeline
An automated flow that collects, cleans, and moves data from sources to destinations (e.g., website to database to reports).
Data privacy
Controlling what customer and business data is collected, where it goes, and who can access it — especially important when AI touches it.
Deep learning
Machine learning using large neural networks with many layers; the approach behind modern language, vision, and speech systems.
Diffusion model
The AI architecture behind modern image generators; it learns by reversing a process of gradually adding noise to images.
Intelligent document processing
AI that reads, extracts, and organizes data from documents like invoices, contracts, and forms.
Inference
When a trained AI model makes predictions on new data. Training builds the model; inference puts it to work.
Integration
A connection between two software systems so data flows between them without manual entry.
Intent recognition
AI figuring out what a user wants from their words (e.g., 'book a table' signals a reservation intent). Core to voice agents and chatbots.
IVR
Interactive voice response — the old 'press 1 for sales' phone menus. Voice AI replaces these with natural conversation.
K
Keywords
The words and phrases customers type into search. Content targets the keywords its buyers actually use.
KPI (key performance indicator)
A measurable value showing how effectively a business or process is achieving its goals.
L
Landing page
A focused web page designed for a single action (sign up, buy, call). Where ad traffic should land, not your homepage.
Large language model (LLM)
An AI trained on massive amounts of text that can understand and generate human language. Examples include GPT, Claude, and Gemini.
Lead nurturing
Staying in touch with not-yet-ready buyers through useful follow-ups until they're ready.
Lead scoring
Ranking leads by likelihood to buy so follow-up effort goes to the best opportunities first.
Local pack (map pack)
The three map listings Google shows for local searches. The most valuable real estate in local SEO.
M
Machine learning (ML)
A branch of AI where software improves at a task by studying examples, instead of following hand-written rules.
Marketing automation
Software that automates marketing tasks: email sequences, lead scoring, social posting, and ad targeting.
Meta description
The short summary under your page title in search results. It doesn't directly affect rankings, but it affects clicks.
Model
The trained AI itself: the file of learned patterns that turns inputs into outputs.
Multimodal AI
AI that understands multiple input types together -- text, images, audio, video -- the way humans do.
N
NAP
Name, Address, Phone number. Keeping it identical everywhere is a local SEO fundamental.
Natural language processing (NLP)
The field of AI focused on understanding and generating human language.
Neural network
A computing model loosely inspired by the brain: layers of simple units that learn patterns from data.
O
OCR (optical character recognition)
Technology that converts images of text (scans, photos) into machine-readable, searchable text.
P
Personalization
Tailoring content, offers, or experiences to individual users based on their data and behavior.
Predictive analytics
Using historical data and AI to forecast future outcomes: sales, churn, demand, or maintenance needs.
Prompt
The instruction you give an AI. Clear, specific prompts get dramatically better results.
Prompt engineering
The practice of writing effective instructions for AI: context, examples, constraints, and output format.
Q
Quantization
Compressing an AI model to run faster with less memory and minimal quality loss. Key to running AI on smaller devices.
R
RAG (retrieval-augmented generation)
Connecting an AI to your documents so it answers from your actual content instead of guessing. The standard fix for hallucinations in business use.
Rate limiting
Restricting how many API requests can be made in a time window, to prevent abuse and control costs.
Recommendation engine
AI that suggests products, content, or actions based on user behavior (the 'customers also bought' box).
ROI
Return on investment — what you get back relative to what you spent. The number every automation project should be measured against.
RPA (robotic process automation)
Bots that mimic human clicks and keystrokes to operate software that has no API. Brittle compared to API-based automation.
S
Schema markup
Structured data — code on your pages that tells Google exactly what the content means (reviews, FAQs, business info), enabling rich search results.
SDK (software development kit)
A package of tools, libraries, and documentation for building on a platform or integrating a service.
Sentiment analysis
AI detecting emotion or opinion in text (positive, negative, neutral). Used for reviews, support tickets, and social monitoring.
SEO
Search engine optimization — making your website and content rank higher in search results.
Sitemap
A file listing your site's pages so search engines can find and index them all.
Speech-to-text (STT)
AI that converts spoken audio into written text. The 'ears' of a voice agent.
T
Text-to-speech (TTS)
AI that converts written text into natural-sounding speech. The 'voice' of a voice agent.
Tokens
The chunks of text (roughly word pieces) that language models read and write. API pricing is usually per token.
Training data
The examples an AI studies during training. Quality and coverage of training data largely determine what the AI can do.
Transformer
The neural network architecture behind modern LLMs. It processes entire sequences at once using a mechanism called attention.
Trigger
The event that starts an automated workflow: a form submission, a missed call, a calendar event.
V
Vector database
Storage designed for embeddings, so an AI can instantly find the most relevant documents for a question.
Voice agent
An AI that handles phone calls or voice interactions: it answers, routes, books appointments, and follows up like a tireless receptionist.
W
Webhook
An automatic notification one system sends another when something happens (e.g., 'new payment received'). The trigger for real-time automation.
Workflow automation
Software that carries out a multi-step business process automatically when triggered (e.g., new lead → CRM entry → follow-up sequence).