Appendix D: Glossary of AI Terms for Trade Professionals
This glossary defines 50 AI and technology terms in plain English, with a trade-specific example for each one. No computer science degree required. If a term comes up in a sales pitch, a conference, or another chapter of this book, you can flip here and get a straight answer about what it means.
AI (Artificial Intelligence): Software that can do things that normally require human thinking — like understanding language, recognizing images, making decisions, and learning from experience. It is not a robot. It is software. Trade example: An AI phone system that listens to a caller describe their plumbing emergency, understands what they need, and books them an appointment — just like a trained receptionist would.
Algorithm: A set of rules or steps that software follows to solve a problem or make a decision. Think of it like a troubleshooting flowchart, but for a computer. Trade example: The algorithm in a scheduling app decides which technician to send to a job based on rules like proximity, skill level, and available parts on their truck.
API (Application Programming Interface): A way for two software systems to talk to each other and share data. It is like a universal adapter that lets different tools plug into each other. Trade example: Your CRM uses an API to send new appointment data to your accounting software automatically, so you do not have to enter it twice.
Automation: Using software to perform tasks that a person used to do manually, without human intervention once it is set up. Trade example: After a job is marked complete, your system automatically sends the invoice, follows up with a review request three days later, and adds the customer to your maintenance reminder list — all without you lifting a finger.
Bias: When an AI system consistently produces unfair or skewed results because of flaws in its training data or design. AI learns from historical data, and if that data reflects human biases, the AI can too. Trade example: If an AI lead-scoring system was trained mostly on data from one neighborhood, it might undervalue leads from other areas — even if those leads are equally good. This is why it is important to review AI recommendations rather than blindly trusting them.
Chatbot: A software program that can have text-based conversations with people, usually on a website or messaging app. Modern chatbots powered by AI are far more capable than the clunky ones from a few years ago. Trade example: A chatbot on your website that answers common questions like "What areas do you serve?" and "What is your service call fee?" at 2 AM when no one is in the office.
Cloud: Storing data and running software on remote servers accessed over the internet, instead of on a computer in your office. When someone says a tool is "cloud-based," it means you access it through a web browser or app, and your data is stored securely online. Trade example: Your field service management software is cloud-based, so your techs can access job details, customer history, and scheduling on their phones from any jobsite.
Computer Vision: AI that can understand and analyze images and video — identifying objects, reading text, detecting damage, or measuring dimensions from a photo. Trade example: AI that analyzes drone photos of a roof and automatically identifies missing shingles, cracked flashing, and hail damage without a human reviewing every image.
CRM (Customer Relationship Management): Software that tracks all your interactions with customers — calls, emails, jobs, invoices, and notes — in one place. An AI-powered CRM adds intelligent features like lead scoring, automated follow-ups, and predictive analytics. Trade example: Your CRM shows that a customer had their AC serviced two years ago, their equipment is 12 years old, and they viewed the "new system" page on your website last week. The AI flags them as a high-probability replacement lead.
Dashboard: A visual display that shows your key business metrics in one place — like the instrument panel of your truck, but for your business data. Trade example: Your morning dashboard shows yesterday's revenue, today's scheduled jobs, outstanding estimates, and online reviews received — all updated in real time.
Data: Any recorded information. In a business context, data includes customer records, job history, financial transactions, phone call logs, website visits, and anything else that can be measured or tracked. Trade example: Every job you have completed, every call you have received, and every invoice you have sent is data. AI tools use this data to find patterns and make your business smarter.
Data Privacy: Rules and practices around how personal information is collected, stored, used, and protected. Regulations like state privacy laws require businesses to handle customer data responsibly. Trade example: When your AI phone system records calls and stores customer information, you need to comply with local recording consent laws and keep that data secure. Reputable AI vendors handle most of this for you, but you should ask about their privacy practices.
Deep Learning: A more advanced form of machine learning that uses layers of artificial "neurons" to find complex patterns in data. It is what powers the most impressive AI capabilities like image recognition and natural language understanding. Trade example: The deep learning model behind a pipe inspection AI has been trained on millions of sewer camera images, so it can identify a hairline crack that even an experienced tech might miss on a small camera screen.
Edge Computing: Processing data on the device itself (like a phone or sensor) rather than sending it to the cloud. This makes AI faster because data does not have to travel to a distant server and back. Trade example: An AI-powered diagnostic tool on a technician's tablet that can analyze equipment readings instantly on-site, even in a basement with no cell signal, because the processing happens on the device itself.
Embedding: A way of representing information (like a word, sentence, or image) as a list of numbers that captures its meaning. This lets AI compare and find similar things. Trade example: When a customer searches your website for "my house is too hot upstairs," the embedding converts that into a meaning representation that matches it with your content about HVAC zoning solutions — even though no one typed those exact words.
Fine-Tuning: Taking an existing AI model and training it further on your specific data so it performs better for your particular use case. Trade example: A general-purpose AI estimating tool is fine-tuned on your company's historical job data, local material prices, and your pricing structure, so its estimates match your actual costs more accurately.
Generative AI: AI that creates new content — text, images, video, code — rather than just analyzing existing content. ChatGPT, image generators, and AI writing tools are all generative AI. Trade example: You describe a job to a generative AI tool, and it writes a professional proposal, creates a social media post about the completed work, and drafts a follow-up email to the customer — all from your brief description.
GPU (Graphics Processing Unit): A computer chip originally designed for video games that turns out to be extremely good at the math required for AI. This is why AI development accelerated so rapidly — the hardware already existed. Trade example: You do not need to know what a GPU is to use AI tools. But when someone says AI requires "a lot of compute power," this is the hardware they are talking about.
Hallucination: When an AI generates information that sounds confident and plausible but is actually wrong or made up. This is one of the known limitations of current AI systems. Trade example: If you ask an AI chatbot a very specific technical question about a particular equipment model's torque specifications, it might generate a convincing but incorrect answer. Always verify critical technical specifications from manufacturer documentation.
Integration: Connecting two or more software systems so they share data and work together seamlessly. Trade example: Your AI phone answering system integrates with your scheduling software, so when the AI books an appointment, it appears on your dispatcher's screen automatically — no manual re-entry needed.
IoT (Internet of Things): Physical devices — sensors, thermostats, monitors, cameras — that connect to the internet and share data. These devices create the data that AI systems analyze. Trade example: A smart thermostat connected to the internet sends temperature and runtime data to an AI system that monitors your customer's HVAC equipment for early signs of failure.
Knowledge Base: An organized collection of information that an AI system can search and reference when answering questions or making decisions. Trade example: You build a knowledge base containing your company's service procedures, pricing guidelines, warranty policies, and FAQs. Your AI phone system and chatbot reference this knowledge base to give accurate answers to customer questions.
Large Language Model (LLM): The type of AI behind tools like ChatGPT and Claude. It is called "large" because it was trained on enormous amounts of text, and "language model" because it understands and generates human language. Trade example: When you type "Write me a follow-up email for a customer who got a furnace quote last week but has not responded," a large language model generates a natural, professional email that sounds like you wrote it.
Lead Scoring: Using data and AI to rank potential customers by how likely they are to buy, so you focus your time on the hottest prospects. Trade example: Your CRM assigns a score of 92 to a homeowner who called about a new roof, lives in a neighborhood where you have done 15 jobs, and has insurance that covers storm damage. It assigns a 34 to someone who filled out a web form at 3 AM asking for a ballpark price. You call the 92 first.
Machine Learning (ML): A type of AI where the software learns from data and improves over time without being explicitly programmed for every scenario. Instead of a developer writing rules for every situation, the system figures out the patterns itself. Trade example: A machine learning system analyzes two years of your dispatch data and learns that jobs in older homes take 20% longer on average. It automatically adjusts time estimates for appointments in pre-1980 construction.
Model: In AI, a model is the trained software that makes predictions or generates output. Think of it as the "brain" that has been educated on data. Trade example: The AI model in your estimating software has been trained on thousands of roofing jobs. When you feed it a roof measurement, it uses what it learned to predict material quantities and labor hours.
Natural Language Processing (NLP): AI's ability to understand, interpret, and respond to human language — whether spoken or written. This is what makes AI phone systems and chatbots possible. Trade example: A customer calls and says, "My AC is making a weird grinding noise and the house is not getting cool." NLP lets the AI understand this is an AC repair call, the symptoms suggest a compressor or fan motor issue, and it should be flagged as urgent.
Neural Network: A software architecture inspired by the human brain, made up of layers of connected nodes that process information. Neural networks are the foundation of most modern AI. Trade example: You do not need to understand neural networks to use AI. Just know that when someone says a tool uses a neural network, it means the AI behind it can learn complex patterns — like recognizing different types of roof damage from photos.
Onboarding: The process of setting up a new software tool and learning how to use it. Good AI tools have guided onboarding that gets you up and running quickly. Trade example: When you sign up for an AI scheduling tool, the onboarding process walks you through importing your customer list, setting your service area, defining your team's skills, and configuring your booking rules.
Opt-In / Opt-Out: Giving or revoking permission for something — usually related to receiving communications or sharing data. Trade example: When your AI system sends automated text messages to customers, those customers need to have opted in to receive texts from you. Sending unsolicited texts can result in legal issues and fines.
Personalization: Tailoring content, recommendations, or experiences to an individual person based on their data and behavior. Trade example: Your AI email system sends different messages to different customers — a maintenance reminder to the homeowner whose furnace is due for a tune-up, a rebate notification to the one who asked about a new AC, and a thank-you note to the one whose job was completed yesterday.
Platform: A software system that serves as a foundation for other tools, applications, or services to build on. Trade example: Your field service management platform is the central system that your scheduling, dispatching, invoicing, and customer communication tools all connect to.
Predictive Analytics: Using historical data and AI to forecast future outcomes — what is likely to happen, when, and to whom. Trade example: Predictive analytics tells you that based on equipment age, usage patterns, and maintenance history, a customer's water heater has a 78% chance of failing within the next 6 months. You reach out with a proactive replacement offer.
Prompt: The instruction, question, or input you give to an AI system. The quality of the output depends heavily on the quality of the prompt. Trade example: Instead of telling ChatGPT "write me an ad," a better prompt is "Write a Facebook ad for an HVAC company in Charlotte, NC. Target homeowners with systems over 10 years old. Emphasize energy savings and include a $50-off tune-up offer. Keep it under 100 words."
Prompt Engineering: The skill of writing effective prompts to get the best possible output from an AI system. It is part art, part science, and it gets easier with practice. Trade example: You learn that when asking AI to write a proposal, including details like "Use a professional but friendly tone, include a warranty section, and keep it under one page" produces much better results than a vague request.
RAG (Retrieval-Augmented Generation): A technique where an AI looks up relevant information from a specific database before generating its response, making the output more accurate and grounded in facts. Trade example: Your AI chatbot uses RAG to pull from your company's actual pricing, service descriptions, and warranty terms when answering customer questions — so it gives accurate information about your business instead of generic answers.
ROI (Return on Investment): The profit you make compared to what you spent. Expressed as a percentage or a multiple. Trade example: You spend $200/month on an AI phone answering service and it generates $3,000/month in additional booked revenue. Your ROI is ($3,000 - $200) / $200 = 1,400%, or a 14x return.
SaaS (Software as a Service): Software you access over the internet and pay for monthly or annually, instead of buying and installing it on your computer. Almost all modern business tools are SaaS. Trade example: Your CRM, scheduling software, and accounting system are all SaaS products. You pay a monthly fee, access them through a web browser or app, and the company handles all the technical maintenance and updates.
Scheduling Optimization: Using AI to create the most efficient schedule possible, considering factors like technician skills, location, traffic, job duration, parts availability, and customer preferences. Trade example: Your AI scheduling system books a furnace repair for your most experienced tech because the unit is a complex model, routes your apprentice team to three straightforward installations that are all in the same neighborhood, and leaves a buffer for the emergency call that statistically comes in every Tuesday afternoon.
Sentiment Analysis: AI that determines the emotional tone of text — whether a review, email, or social media post is positive, negative, or neutral. Trade example: Your AI reputation management tool scans every new online review. When it detects a negative review, it immediately alerts you and drafts a professional response for your approval.
SEO (Search Engine Optimization): Techniques to make your website appear higher in Google search results when potential customers search for your services. Trade example: AI-powered SEO tools analyze your website and tell you that adding content about "emergency AC repair in [your city]" and "water heater replacement cost" would help you rank for searches that local homeowners are actually making.
Smart Home: A home equipped with internet-connected devices — thermostats, locks, cameras, lights, sensors — that can be controlled remotely and automated. Trade example: As an electrician or HVAC contractor, smart home installations are a growing revenue stream. Understanding the technology and being able to install, configure, and troubleshoot smart devices sets you apart from competitors who avoid them.
Subscription Model: A business model where customers pay a recurring fee (monthly or annually) for ongoing service, rather than paying per transaction. Trade example: Instead of charging customers $150 for an annual furnace tune-up, you offer a $19/month comfort plan that includes the annual tune-up, priority scheduling, a 15% repair discount, and no after-hours surcharges. You get predictable recurring revenue; they get peace of mind.
Supervised Learning: A type of machine learning where the AI learns from labeled examples — data where the correct answer is already known. Trade example: A pipe inspection AI learned to identify root intrusion by studying thousands of sewer camera images that were labeled by experienced plumbers as "root intrusion" or "no root intrusion." The more labeled examples it studied, the more accurate it became.
Token: The basic unit of text that AI language models process. Roughly, one token equals about three-quarters of a word. AI services often charge based on tokens processed. Trade example: When you use ChatGPT to write a 500-word proposal, it processes approximately 670 tokens of your input and generates approximately 670 tokens of output. Most pricing plans include thousands of tokens per dollar, so individual requests cost fractions of a cent.
Training Data: The information used to teach an AI model. The quality and quantity of training data directly determines how good the AI performs. Trade example: An AI estimating tool trained on 100,000 real-world roofing jobs will produce more accurate estimates than one trained on 500 jobs. This is why AI tools ask you to import your historical data — they get better when they learn from your specific business.
Transfer Learning: Using knowledge an AI learned from one task to help it perform better on a related task, even with less training data. Trade example: An AI that learned to analyze commercial roof images can transfer that knowledge to residential roofs. It already understands what shingles, flashing, and damage look like — it just needs a smaller set of residential examples to adapt.
UI (User Interface): The screens, buttons, menus, and visual elements you interact with when using software. A good UI makes a tool easy to use; a bad UI makes it frustrating. Trade example: Your dispatcher needs a UI that shows today's schedule, tech locations, and unassigned jobs in a single glance — not one that requires clicking through five screens to see basic information.
Unsupervised Learning: A type of machine learning where the AI finds patterns in data without being told what to look for. The system discovers groupings and relationships on its own. Trade example: An unsupervised learning system analyzes your customer database and discovers that customers in a certain age of home, in a certain neighborhood, with a certain equipment brand tend to need replacements around the same time. You did not tell it to look for this pattern — it found it.
Uptime: The percentage of time a software service is available and working. Industry standard for good uptime is 99.9% or higher. Trade example: If your AI phone answering system has 99.9% uptime, that means it is down for less than 9 hours per year. At 99% uptime, it is down for 87 hours — more than 3.5 full days. That difference matters when every missed call is a potential job.
Voice AI: AI that can understand spoken language and respond with natural-sounding speech. This is the technology behind AI phone answering systems and voice assistants. Trade example: A customer calls your number. Voice AI answers in a natural, friendly voice, understands their request for an AC repair, asks qualification questions, and books an appointment — all through a normal phone conversation.
Webhook: An automated notification sent from one software system to another when something specific happens. It is like a digital tap on the shoulder. Trade example: When a customer submits a form on your website, a webhook instantly sends that lead information to your CRM, triggers an automated text response to the customer, and creates a task for your sales team — all within seconds.
Workflow: A defined sequence of steps that a task follows from start to finish. AI-powered workflow automation moves tasks through these steps with minimal human intervention. Trade example: Your job workflow goes: lead received, estimate created, estimate approved, job scheduled, job completed, invoice sent, payment received, review requested. AI can automate the transitions between most of these steps.
Zero-Shot Learning: AI that can perform a task it was not specifically trained on, by using its general knowledge to figure out what is needed. Trade example: You ask an AI to write a proposal for a radiant floor heating installation, even though it was never specifically trained on that service type. Because it understands proposals, heating systems, and construction in general, it produces a reasonable first draft that you then customize.
Quick Reference Card
Here are the 10 terms you will encounter most often as you evaluate and adopt AI tools:
| Term | One-Line Definition |
|---|---|
| AI | Software that thinks and learns |
| Automation | Tasks that run themselves |
| CRM | Your customer database with superpowers |
| Cloud | Software you access online, not installed locally |
| Integration | Making two tools talk to each other |
| LLM | The AI brain behind ChatGPT and similar tools |
| Prompt | The instruction you give to AI |
| SaaS | Software you rent monthly |
| API | The connector between software systems |
| ROI | Whether the investment is paying off |
When a sales rep throws jargon at you, flip to this glossary. Understanding the language gives you the confidence to ask the right questions and make informed decisions about the AI tools that will shape the future of your business.