Plain-language definitions
Glossary
The AI terms a CEO will run into this year. Each entry is a 90-second read · what it is, why it matters, and where you'll see it.
A
- AI agent An AI model put to work in a loop: it decides, acts with a tool, looks at the result, and repeats until the job is done.
- AI assistant A reactive, human-in-the-loop helper: you ask, it responds, and you steer the next move.
- AR/AP aging Two reports that show who owes you money and who you owe, grouped by how overdue each amount is. AR aging is money coming in (customer invoices). AP aging is money going out (bills you have to pay).…
C
- cash vs accrual The two ways to keep your books. Cash basis records money when it actually moves. Accrual basis records it when it is earned or owed, regardless of when the cash arrives. Which one you use changes…
- Claude Code Anthropic's command-line tool for coding with Claude. Runs in your terminal, edits files in your project, and remembers context across sessions.
- Claude Routine A workflow you've handed to Claude Code to run on a schedule, on its own, without you in the loop. The difference between asking for help and owning a process that runs while the desk is empty.
- context window The slice of text a model can see at one moment: its working memory, measured in tokens.
D
- DPA Data processing agreement: the contract that spells out how a vendor handles your data on your behalf. What they can do with it, how long they keep it, whether they train on it, and what happens if…
F
- frontier model One of the handful of most capable AI models at the leading edge at any given moment, built by the few labs that can afford to train them.
G
- general ledger The master record of every financial transaction in your business, sorted by account. Your P&L and balance sheet are both built from it. When people say "the books," the general ledger is what they…
- Granola A meeting-notes app that records your calls, transcribes them, and writes a clean summary you can actually search later. The raw material every meeting-driven workflow starts from.
- Granola Recipe A saved, repeatable prompt in Granola that runs the same way against every meeting. Instead of re-typing "pull out the action items and decisions" after each call, you build the instruction once as…
H
- hallucination When an AI states something false as if it were true, fluently and confidently, because it predicts plausible text rather than checking facts.
P
- PII Personally identifiable information: any data that can identify a specific person. Names, emails, phone numbers, addresses, Social Security numbers, customer records, employee files. The category of…
- Prompt The full instruction you give an AI model. Not just a question. A prompt is the complete package: task, context, format, and constraints. The model works only with what you hand it, so the quality…
R
- red team The practice of deliberately assigning someone (or something) the job of attacking your plan to find its weak points before reality does. The term comes from military and security work. With AI, you…
Z
- zero data retention A contractual setting where the AI provider keeps none of your inputs or outputs after processing them. Nothing is stored, nothing is logged, nothing can be handed over later. The strongest data…