ProductFeaturesAgentsSecurityPricingIntegrationsCompareBook a demo
Learn

The aanty glossary

38 plain-language explanations of the concepts behind an AI-native communication platform — from typed messages and agents to row-level security and pricing — each with a note on how aanty actually implements it.

Pick a term

Concept · AI · Security · Pricing

Every entry is a short, honest definition first, then how aanty relates to it — no invented facts or metrics.

Concept

Typed Messages

A typed message is a message that carries a structured kind — question, decision, task, approval, and so on — instead of being an undifferentiated blob of text. The type drives how the system routes, prioritizes and searches it.

Concept

Conversation Graph

A conversation graph models communication as a connected structure of messages, threads, entities and people — instead of a flat list of channels — so any two related things can be linked, queried and traced.

Concept

The Briefing

A briefing, in this context, is an AI-composed summary that answers "what happened, what needs me, and what can I ignore?" — designed to be the default home screen instead of a channel list.

Concept

Needs-You Queue

A needs-you queue is a short, trusted list of items that genuinely require a person's attention, each with a stated reason — as opposed to an unread count that just measures how much you haven't looked at yet.

Concept

Attention Economics

Attention economics describes how a product's business incentives shape what it shows you — engagement-funded tools tend to maximize time-in-app and unread counts, while tools funded another way can optimize for less reading instead.

Concept

Notification Fatigue

Notification fatigue is the desensitization and stress that comes from receiving more alerts than a person can meaningfully act on — the point where notifications stop directing attention and start just adding noise.

AI

Message Classification

Message classification is the process — usually AI-assisted — of assigning a category or type to a message after it's written, so downstream systems can route, prioritize and search it without a human tagging it by hand.

Concept

Decision Ledger

A decision ledger is a searchable log of decisions an organization has made, each with its context, the option chosen, an owner, and — often — a date to revisit it, so "why did we choose X?" has a direct answer instead of a scroll through chat history.

AI

Question → Answer Matching

Question-answer matching checks a newly asked question against a store of previously answered ones before routing it to a person — so a question that's already been answered gets the existing answer instead of asking someone to type it again.

AI

Model Context Protocol (MCP)

MCP is an open protocol, now under Linux Foundation governance, that standardizes how AI models and agents connect to external tools and data sources — an app can expose an "MCP server" that any compliant AI client can query.

AI

AI Agents

An AI agent is software that can perceive context, decide on actions, and carry them out — sending messages, calling tools, taking real steps — rather than just answering a single prompt with text.

AI

Agents as Principals

"Agents as principals" means treating AI agents as first-class identities in a system — with their own identity, permissions and audit trail — the same primitives a human user has, rather than a special case bolted onto the user model.

AI

Capability Scoping

Capability scoping means limiting what an actor — especially an AI agent — is allowed to do to a specific, named set of tools and permissions, instead of granting broad or default access that gets restricted after the fact.

AI

Agent Provenance

Agent provenance is the record attached to an AI agent's output showing where it came from — which model produced it, its confidence, and what sources it cited — so a human can verify or challenge it instead of taking it on faith.

Security

Row-Level Security (RLS)

Row-level security is a database feature — built into Postgres — that restricts which rows a query can see or modify based on a policy, so tenant isolation is enforced by the database itself rather than only by application code.

Security

FORCE Row-Level Security

FORCE ROW LEVEL SECURITY is a Postgres setting that makes a row-level-security policy apply even to the table's owner and other privileged roles that would otherwise bypass it by default — closing a common isolation gap.

Security

ReBAC (Relationship-Based Access Control)

ReBAC grants access based on relationships between entities — "is a member of," "is the guardian of," "is a client of" — rather than fixed roles, letting permission checks follow the actual structure of who is connected to what.

Security

RBAC (Role-Based Access Control)

RBAC grants permissions based on a user's assigned role — admin, member, moderator — rather than to the user individually, so access can be managed by changing role membership instead of editing per-user permissions.

Security

End-to-End Encryption (E2EE)

End-to-end encryption means content is encrypted on the sender's device and only decrypted on the recipient's device, so the service operating the platform in between never has access to the plaintext — even if its own servers are compromised.

Security

Append-Only Audit Log

An append-only audit log is a record of events that can only be added to, never edited or deleted — enforced at the database or storage layer — so the history of what happened is provable rather than just claimed.

Concept

Event Sourcing

Event sourcing is an architecture where a system stores the sequence of events that happened — not just the current state — and derives current state by replaying or projecting those events, so history is the source of truth.

Concept

Entity-Linked Threads

An entity-linked thread is a conversation thread that's attached to a specific external object — an invoice, a task, a risk, a customer record — so the discussion and the object it's about live in one place instead of being disconnected.

Security

Permission-Trimmed Search

Permission-trimmed search restricts search results to what the searching user is actually allowed to see, applying access checks before ranking rather than filtering an already-ranked list afterward — so nothing unauthorized is ever scored, shown, or briefly visible.

AI

Org Memory

Org memory is a knowledge layer built automatically from an organization's conversations — profiles of projects, customers and systems that accumulate context over time — visible and editable rather than a hidden model.

Concept

Quiet Hours

Quiet hours are a scheduled, timezone-aware window during which non-urgent notifications are held back rather than delivered — with an explicit exception for genuinely critical alerts that are allowed to break through.

Concept

Priority Routing

Priority routing means delivering notifications based on how urgent or important a message actually is, rather than treating every message from a subscribed channel identically — so a critical alert and a routine update don't compete for the same attention.

Pricing

Per-Seat Pricing

Per-seat pricing charges a fixed amount for every licensed user (seat), regardless of how much they actually use the product — the dominant SaaS pricing model for team collaboration tools, and a recurring source of customer frustration as prices rise.

Pricing

Unlimited Message History

Unlimited message history means every past message stays searchable and accessible indefinitely, on every pricing tier — as opposed to free or lower tiers that only retain a limited window (commonly 90 days) before older messages become inaccessible.

Concept

Client Portal

A client portal is a branded, purpose-built external surface where a client or guest sees only the threads, files and approvals relevant to them — a dedicated experience, not a stripped-down version of the internal tool.

Security

Sensitive Community Mode

Sensitive community mode is a single named bundle of safety settings — pseudonymity, human-routed crisis escalation, mandatory moderation, stricter audit — applied together as one toggle, for communities serving vulnerable populations.

Security

Pseudonymity

Pseudonymity means a person's visible, in-product display identity is deliberately decoupled from their real billing, authentication or legal identity — so other participants see a consistent handle, not a name that traces back to who they really are.

Security

SCIM Provisioning

SCIM (System for Cross-domain Identity Management) is a standard protocol for automatically syncing user accounts — creation, updates, deactivation — from an identity provider to connected applications, instead of managing accounts by hand in each one.

Security

SSO (OIDC / SAML)

Single sign-on lets a person authenticate once with their organization's identity provider and use that same session across multiple applications — OIDC and SAML are the two dominant standard protocols that make this interoperable.

Security

Data Loss Prevention (DLP)

Data loss prevention refers to tools and rules that detect and block sensitive information — credentials, personal data, financial identifiers — from being shared inappropriately, whether accidentally or maliciously.

Security

Data Residency

Data residency refers to controlling the specific geographic location where an organization's data is stored and processed — often required by regional privacy law or a customer's own regulatory or contractual obligations.

Security

BYOC (Bring Your Own Cloud)

BYOC is a deployment model where a vendor's software runs inside a customer's own cloud account — AWS, Azure, GCP — rather than the vendor's infrastructure, giving the customer direct control over the environment while the vendor still manages the software.

Security

Self-Hosting

Self-hosting means running a product's software entirely on infrastructure the customer owns and operates — including deployment, updates, and maintenance — rather than relying on the vendor's managed service at all.

AI

Workslop

Workslop is a term for AI-generated work output that looks complete and polished but lacks real substance — content that appears finished on the surface while actually pushing the effort of making it useful onto whoever receives it.

Book a demo

Turn a definition into a demo.

Any of these concepts is easier to see than to read about. Bring a real channel and we'll show the ones that matter to you, live.