Plain-English definitions of terms you'll hear from vendors, in our sessions, and in your AI blueprint.
A to C
10DLC. The US registration system for businesses that send texts from regular 10-digit phone numbers. Most business texting tools require it.
AI (artificial intelligence). Software that can do tasks that normally take human judgment, like understanding language, recognizing patterns, or making simple decisions.
AI agent. An AI system that can take actions toward a goal, such as booking an appointment or updating a record, not just answer questions. See AI Agents vs. Chatbots vs. Automation.
AI assistant. A general term for an AI tool that helps with tasks by chat, voice, or text.
API (application programming interface). A way for two pieces of software to talk to each other directly. When a vendor says "we have an API," it means other tools can connect to it.
Automation. Software that runs a task automatically when something happens, following set rules.
Baseline. Your numbers before you change anything. You need one to measure whether a change helped.
CAN-SPAM. The US law that sets rules for commercial email, including identifying yourself, including a physical address, and honoring unsubscribe requests.
Chatbot. Software that holds a conversation by text or chat. Older ones follow scripts; AI chatbots understand free-form questions.
Confidence score. A number some AI tools give to show how sure they are about an answer. Low-confidence results should go to a person.
Context window. How much text an AI model can consider at once, including your instructions, documents, and the conversation so far.
CRM (customer relationship management). Software that keeps track of your customers, leads, and every interaction with them.
D to H
Drip campaign. A series of emails or messages sent automatically on a schedule.
Fine-tuning. Training an existing AI model further on specific examples for a particular task.
Generative AI. AI that creates new content, such as text, images, or audio, rather than only sorting or analyzing.
Guardrails. Rules and limits that keep an AI system from doing things it shouldn't, like quoting unapproved prices.
Hallucination. When an AI model states something false with confidence. The main reason to keep a human checking important output.
Human in the loop. A setup where a person reviews or approves AI output before it takes effect.
I to L
Integration. A connection between two software tools so information moves between them automatically.
Intent. What a customer is trying to do, like book, complain, or ask a question. AI tools often sort messages by intent.
IVR (interactive voice response). The phone menu system that says "press 1 for sales." AI voice assistants are a newer alternative.
KPI (key performance indicator). A number you track to see how something is going, like days to get paid or missed calls per week.
Knowledge base. A collection of articles that answer common questions. AI assistants can use one to answer customers accurately.
Large language model (LLM). The kind of AI behind tools like Claude and ChatGPT. Trained on large amounts of text to understand and generate language.
Lead scoring. Ranking leads by how likely they are to buy, so you call the best ones first.
M to P
Machine learning. A way of building software that learns patterns from examples instead of following only hand-written rules.
Model. The trained AI system itself. Different models have different strengths, speeds, and costs.
No-code / low-code. Tools that let you build automations by clicking and configuring instead of writing code.
OCR (optical character recognition). Technology that turns images of text, like scanned documents, into editable text.
Opt-in / opt-out. Opt-in means someone actively agreed to receive something. Opt-out means they asked to stop.
Pilot. A small, time-limited test of a tool or process before committing.
Prompt. The instructions you give an AI model.
R to S
RAG (retrieval-augmented generation). A technique where an AI looks up information in your documents before answering, so its answers are based on your actual information.
ROI (return on investment). What you gained compared with what you spent. See How to Measure AI ROI.
Review gating. Asking only happy customers for reviews. It's against Google's policies; ask everyone.
Sentiment analysis. Using AI to judge whether a message or review is positive, negative, or neutral.
SOP (standard operating procedure). A written, step-by-step description of how a task gets done. See Build Your First SOP for AI.
Speech-to-text. Technology that turns spoken audio into written text, used for call transcripts and voice assistants.
T to Z
Text-back. An automatic text sent to someone after a missed call.
Token. A small chunk of text, roughly part of a word, that AI models read and write. Many AI services charge by the token.
Training data. The information an AI model learned from. Ask vendors whether your data is used as training data.
Trigger. The event that starts an automation, like a form being submitted.
Voice agent. An AI system that talks with callers by phone.
Webhook. A way for one app to instantly notify another when something happens.
Workflow. A sequence of steps that gets a job done, manual or automated.
Integration platform. A no-code service that connects your apps and runs automations between them.
Missing a term?
Email hello@aiventureholdings.com and we'll consider adding it.