Automation Terms A-Z
Master automation in marketing: From Workflow Automation to RPA to trigger-based campaigns – all terms for efficient marketing processes.
A
Agent Liability Gap
The agent liability gap is the situation in which harm caused by an AI agent cannot be clearly attributed to any single party, because model provider, tool operator, integrator and deploying company each carry only partial responsibility.
Agent Orchestration
Coordination and control of multiple AI agents to execute complex workflows, including task distribution, communication, and error handling.
Agent Payment Rails
Agent payment rails are the technical and contractual foundation that lets an AI agent trigger payments on behalf of a person or company: delegated mandates, agent-specific payment credentials, amount limits and machine-readable receipts.
Agent Sprawl
Agent sprawl describes a state in which more autonomous or semi-autonomous AI agents run in production than are centrally inventoried, monitored and owned. Each individual agent may be useful, but together they create redundancy, unclear data access and unattributable cost.
Agent Swarm
An agent swarm is a group of homogeneous or heterogeneous AI agents that pursue the same goal in parallel (e.g. 100 copy variants) and evaluate and filter each other.
Agentic CRM
An agentic CRM is a customer relationship system in which AI agents autonomously perform defined tasks along the sales process. Agents work with read and write access to CRM data, use external sources and route critical steps to human approval.
AgentOps
AgentOps bundles practices, tools and roles for the operation, monitoring, cost control, security and continuous improvement of AI agents in production — the logical evolution of MLOps and LLMOps.
AI Orchestration
The coordinated control and integration of multiple AI models, agents, and tools to execute complex, multi-step tasks in an automated workflow.
Ambient Agents
Ambient agents run continuously in the background, watching data sources, inboxes and tools, and proactively suggest or execute actions without an explicit human trigger.
Autonomous Crash Watching & Debugging
Autonomous crash watching and debugging refers to AI agents that automatically observe runtime errors and crash reports, analyse the related logs and code sections and independently produce a diagnosis, an error report or a fix proposal.
Autonomy Level
Autonomy level classifies how much decision-making an agent handles — from L0 (suggest only) via L2 (human-in-the-loop approval) to L5 (fully autonomous with post-hoc review).
B
Botsitting
Botsitting describes the hidden human work required to supervise AI systems — filling in missing context, checking outputs, debugging mistakes, rewriting prompts and dealing with hallucinations. 2026 studies show knowledge workers spend nearly a full working day per week on botsitting.
Browserbase
Browserbase is a platform for running and controlling headless browsers in the cloud. Sessions retain login state, can be routed through proxies, recorded, and driven with natural-language instructions via its own framework Stagehand.
C
Composio
Composio is an integration platform for AI agents that provides tools for numerous SaaS applications, manages per-user authentication and can be connected via MCP among others.
Computer Use
The ability of AI models to operate computers like humans – interpret screenshots, control mouse and keyboard, navigate through interfaces.
Copilot Fatigue
Copilot fatigue describes the exhaustion of employees managing too many AI suggestions, alerts and assistants — with declining usage and worse decision quality.
E
Eval Framework
Systematic framework for evaluating LLM outputs against defined criteria like correctness, relevance, safety, and style.
Eval-Driven Marketing
Eval-driven marketing describes a way of working in which AI-powered marketing functions are checked against a versioned collection of test cases and quality criteria. Changes to prompts, models or knowledge sources are only released once evaluation passes.
H
Human-in-the-Loop
Design principle where humans are involved at critical points in automated AI workflows to validate, correct, or approve decisions.
Human-in-the-Loop Fatigue
Human-in-the-loop fatigue describes the exhaustion and quality decline that happens when humans continuously approve, correct or prioritise AI outputs. After a short time review depth collapses and errors and hallucinations slip through.
Human-in-the-Loop Tax
The human-in-the-loop tax is the total human effort for approval, review, correction and escalation that an AI-assisted process creates, measured against the processing time it saves.
M
Marketing Automation
The use of software to automate repetitive marketing tasks like email campaigns, social media posts, or lead nurturing.
Multi-Agent Orchestration
Multi-agent orchestration coordinates several specialised AI agents (planner, researcher, writer, reviewer) into a workflow — with clear roles, handoffs, budgets and guardrails.
O
One-Shot Code Translation
One-shot code translation refers to the agentic transfer of a working codebase into another language, preserving behaviour and interface, with a deliberately chosen target environment — such as a dependency-free binary. Evaluation runs on tests against the original and measured performance figures, not on the agent's impression.
Oversight Tax
Oversight tax is the time, staffing and process cost of human oversight over AI agents — reviews, approvals, audits, documentation. Analogous to compliance cost in regulated industries.
Q
QA (Quality Assurance)
Quality assurance is the systematic process of ensuring outputs meet defined standards—correctness, consistency, safety, usability, and compliance.
QBR (Quarterly Business Review)
A QBR is a structured quarterly review between a vendor/team and stakeholders to assess performance, outcomes, risks, roadmap, and priorities.
Quality Drift
Quality drift is a gradual degradation of output quality over time due to changes in data, prompts, retrieval corpora, user behavior, or system dependencies.
Quality Filter
A quality filter is a rule or model that blocks, flags, or degrades outputs that fail quality criteria.
Quality Gates
Quality gates are automated (and sometimes human) checks that content or system changes must pass before release or publication.
Quick Fix
A quick fix is a small, fast change intended to mitigate an issue immediately (often a tactical patch), usually followed by a deeper root-cause fix.
Quiet Period
A quiet period is a defined time window where teams avoid making changes that could confound measurement or increase risk.
S
Shadow Agent
Shadow agents are autonomous or semi-autonomous AI agents run outside official governance, often with personal API keys or delegated user permissions, whose actions do not appear in central monitoring.
Spec-Driven Development
Spec-driven development is a way of working in which requirements, acceptance criteria and constraints are captured in a machine-readable specification from which AI agents derive implementation and tests. Changes are made to the specification, not to the generated code.
Sub-Agents & Agent Harnesses
Sub-agents are specialised AI agents with their own context and a bounded mandate, commissioned by a lead agent. The agent harness is the orchestration layer around them: it slices tasks, launches sub-agents in parallel, monitors their runtime and cost and merges partial results into a single outcome.
T
Temporal
Temporal is an open-source durable execution platform, originating from the Cadence project. It orchestrates long-running workflows and individual activities resiliently, with automatic retries, durable state and support for multiple programming languages.
Terminal Multi-Agent Orchestration
Terminal multi-agent orchestration describes a layer above classic terminal multiplexers such as tmux that starts, names, monitors and bundles several concurrently running agent sessions. Its essential contribution is not launching but notifying: it detects when a session is waiting for human input and draws attention to it.
Trigger.dev
Trigger.dev is an open-source platform for background tasks and workflows in TypeScript that run without timeouts, support retries and waits, and can be self-hosted.