AI agent

An AI agent is a system that uses a language model to pursue a goal by planning steps and using tools.

Written by Reflective Data experts · updated Oct 5, 2026

Instead of only answering questions, an agent can search, query a warehouse, call APIs or draft emails, then decide what to do next. The quality of the tools, permissions and checks around it largely determines how useful and safe it is.

Agents need reliable, governed data to give reliable answers.

Agent harness AI guardrails Large language model (LLM)Model Context Protocol (MCP)

// how we can help

We build company brains, agent harnesses and AI bots on top of data you can trust. Custom AI harnesses → · The data platform behind it

Reranking

Reranking reorders retrieved results with a more precise model, so the best passages go to the…

human-in-the-loop.md

Human in the loop

Human in the loop means a person reviews or approves an AI output or action before…

token.md

Token

A token is a small piece of text, often part of a word, that language models…

embeddings.md

Embeddings

Embeddings are numerical representations of text or other data that capture meaning, so similar things are…

chatgpt.md

ChatGPT

ChatGPT is OpenAI conversational AI assistant, built on its GPT family of models.

agentic-workflow.md

Agentic workflow

An agentic workflow is a process in which an AI system plans steps, uses tools and…

← full analytics dictionary