AI Agent Glossary

Software that uses AI to take autonomous actions on behalf of a user, including planning multi-step tasks and executing workflows.

Large Language Model - a neural network trained on vast text data that can understand and generate human-like language. Examples: GPT-4, Claude, Gemini.

Retrieval-Augmented Generation - retrieves relevant information from a knowledge base before generating a response, grounding the AI in factual data.

Model Context Protocol - an open protocol standardizing how AI models connect to external tools and data sources.

Designing and refining prompts to get better, more consistent outputs from AI models.

Further training a pre-trained AI model on specific data to improve performance on a particular use case.

Using software to execute tasks automatically based on triggers and conditions. AI agents add intelligence by handling decisions.

Platforms that allow building automations using visual interfaces instead of writing code.

Application Programming Interface - rules allowing software applications to communicate. How AI agents connect to tools and data.

When an AI generates plausible but factually incorrect information. RAG and grounding techniques mitigate this.

A workflow where certain decisions require human approval before the AI proceeds.

Constraints and rules on AI agents to prevent undesired behavior, limiting what agents can do, say, or access.

Basic unit of text LLMs process - roughly 4 characters. Token counts determine API costs and context limits.

Maximum text an AI model can process at once, measured in tokens. Larger windows handle longer documents.

Specialized database storing data as mathematical vectors, enabling AI to search by meaning rather than keywords.

A workflow where AI agents make decisions and choose tools autonomously rather than following a fixed path.

Architecture where multiple specialized AI agents collaborate, each handling a specific domain.

Return on investment from automation: time saved and errors reduced minus tool and setup costs.

Using AI to rank potential customers by their likelihood to buy based on behavioral and firmographic signals.

Using AI to identify customers likely to cancel. Agents act on churn signals with retention workflows.