All technical terms explained simply.
Think-Act-Observe reasoning loop at the core of every AI agent's operation.
Autonomous program that reasons, uses tools, and adapts based on results to accomplish a task.
Interface that allows two pieces of software to communicate with each other (e.g., querying a weather service).
Rule that connects an exchange to a queue, defining which messages the queue should receive.
Pattern defined in a binding, against which the exchange compares incoming message routing keys.
Service or application that retrieves and processes messages from a queue.
Exchange type that routes messages only to queues whose binding key exactly matches the routing key.
RabbitMQ component that receives messages from producers and routes them to the right queues based on rules.
Exchange type that broadcasts every message to all bound queues, ignoring the routing key.
Language model capable of understanding and generating text. It's the "brain" of an AI agent.
Software intermediary that receives messages from one service and distributes them to the right recipients based on routing rules.
AI program trained to accomplish a specific task, such as understanding text, generating images, or transcribing audio.
Coordination of an agent's different components (LLM, tools, memory) to accomplish a task.
Simple technique that associates keywords with predefined responses.
Structured chain of processing steps, where the output of each step feeds the next one.
Service or application that sends messages to an exchange in a messaging system.
Instruction or question sent to a language model to obtain a response.
Architecture pattern where producers publish messages without knowing who will receive them, and consumers subscribe to messages they're interested in.
Storage buffer where messages wait to be retrieved and processed by a consumer.
Label attached to a message by the producer, used by the exchange to decide routing.
Execution environment that runs the agent, manages tool calls, and orchestrates the agent loop.
Technique that displays the response progressively, word by word, instead of waiting for the complete response.
Hidden instructions given to the agent by its designer to guide its behavior, tone, and constraints.
External function that an AI agent can call to interact with the outside world (API, database, etc.).
Mechanism by which an AI agent selects and calls an external function to accomplish a task.
Exchange type that routes messages by comparing the routing key to patterns with wildcards (* and #).