Glossary¶
Key terms and concepts used throughout Functional AI.
Quick Navigation
Use your browser's search (Ctrl+F / Cmd+F) to find specific terms quickly.
A¶
Assistant¶
An AI chatbot that you create and configure in Functional AI. Each assistant has its own instructions, personality, and connected resources.
AUTO_PAY¶
An overflow handling mode where messages beyond your monthly quota are automatically charged at per-message rates, allowing uninterrupted service.
Avatar¶
The profile image displayed alongside messages in chat. Both assistants and users can have custom avatars.
C¶
Channel¶
A platform or medium through which users interact with your assistant. Examples: website widget, Instagram, Messenger, Shopify.
Channel Rules¶
Automation settings that control how your assistant behaves on each channel, including auto-reply settings, delays, and escalation triggers.
Conversation¶
A complete message exchange between a user and your assistant, possibly spanning multiple messages across one or more sessions.
E¶
Embedding¶
A numerical representation of text content that enables semantic search. When you upload documents, they're converted to embeddings.
Escalation¶
The process of flagging a conversation for human review, typically triggered by specific keywords, sentiment, or user requests.
F¶
Function¶
A specific capability that an assistant can invoke, such as searching products or checking order status. Functions are grouped into templates.
Function Template¶
A collection of functions that can be connected to assistants to extend their capabilities. Can be system-provided or custom-created.
I¶
Instructions¶
The core configuration of an assistant that defines its personality, behavior, knowledge scope, and response style. Think of it as the assistant's "job description."
K¶
Knowledge Base¶
See "Store." A collection of documents that provide context for your assistant's responses.
M¶
Message¶
A single exchange in a conversation. Can be from the user or the assistant. Messages count toward your monthly quota.
Message Quota¶
The maximum number of AI responses your account can generate per month, based on your subscription tier.
O¶
Overflow¶
Usage beyond your tier's monthly message quota. Handled according to your overflow settings (STOP or AUTO_PAY).
P¶
Privacy Setting¶
Controls whether an assistant is private (only you can access) or public (shareable via links and embeds).
R¶
RAG (Retrieval-Augmented Generation)¶
A technique where the AI retrieves relevant information from your knowledge bases to enhance its responses with accurate, specific information.
Rate Limit¶
Restrictions on how many requests can be made in a given time period. Prevents abuse and ensures fair usage.
S¶
SafeGuard¶
An optional setting that adds additional safety constraints to assistant responses, helping prevent inappropriate content.
Semantic Search¶
Finding content based on meaning rather than exact keyword matches. Allows questions like "What's the cost?" to find information about "pricing."
Share Link¶
A public URL that allows anyone to chat with your assistant without logging in.
Share Token¶
A unique identifier embedded in share links and widget code that connects to your specific assistant.
Store¶
A knowledge base (vector store) containing uploaded documents. Your assistant searches stores to provide informed responses.
STOP Mode¶
An overflow handling mode where message processing halts when you reach your monthly quota, preventing additional charges.
T¶
Temperature¶
A setting (0.0-1.5) controlling the creativity/randomness of AI responses. Lower values produce consistent responses; higher values produce more varied responses.
Thread¶
A single conversation stream within the Chat interface. Each thread maintains its own context and history.
Tier¶
Your subscription level (Basic, Medium, or Pro), which determines your resource limits and available features.
Token¶
- Share Token: Identifier for your public assistant
- API Token: Authentication credential for programmatic access
- AI Tokens: Units measuring AI input/output length
U¶
Use Case¶
A setting specifying an assistant's domain focus. Options are BASIC (general purpose) or ECOMMERCE (optimized for shopping assistance).
Usage¶
Consumption of platform resources, particularly messages, which count toward your monthly quota.
V¶
Variable¶
A stored configuration value (like API keys or settings) that can be used by functions. Can be regular (visible) or secure (encrypted).
Vector Store¶
The technical name for a knowledge base. Stores document embeddings for semantic search. See "Store."
W¶
Webhook¶
An automated message sent from one system to another when an event occurs. Used for real-time integrations with platforms like Instagram and Messenger.
Widget¶
The chat interface that appears on websites when you embed your assistant. Appears as a floating button and expandable chat window.
Common Abbreviations¶
| Abbreviation | Meaning |
|---|---|
| AI | Artificial Intelligence |
| API | Application Programming Interface |
| DM | Direct Message |
| FAQ | Frequently Asked Questions |
| LLM | Large Language Model |
| OAuth | Open Authorization |
| RAG | Retrieval-Augmented Generation |
| UI | User Interface |
| URL | Uniform Resource Locator |