DE EN

AI Glossary

AI terms, clearly explained.

Technical and business-oriented — from RAG and embeddings to Forward Deployed Engineer and the EU AI Act.

A

Agent (Agentic Workflow)

AI systems that autonomously plan and execute tasks.

Agentic AI

AI systems that pursue goals, plan sub-steps, and use tools to complete them.

AI Governance

Policies and processes for responsible, compliant AI use.

Artificial Intelligence (AI)

Software that performs tasks that normally require human thinking.

C

Chunking

Breaking large documents into smaller, meaningful sections.

Clawdbot / Moltbot / OpenClaw

Terms related to the development and positioning of the OpenClaw framework.

E

Embeddings

Numerical representations of text that capture semantic meaning.

Evaluation / Evals

Systematically measuring and assessing AI quality.

F

Forward Deployed Engineer (FDE)

An experienced software engineer who works directly within a client's context — as part of the team, not as an external advisor. Delivers code, not slides, and takes ownership of architecture decisions.

Fractional CAIO (Chief AI Officer)

An experienced AI and tech leader who works part-time for multiple companies simultaneously — with real strategic ownership, but without a full-time salary.

G

GDPR / Data Privacy (in AI)

Data protection requirements when using AI systems.

Generative AI (GenAI)

AI that creates new content — text, images, code, and more.

Grounding

Anchoring AI responses to real, verifiable sources.

H

Hallucination

When AI generates confidently-stated but factually incorrect information.

Human-in-the-Loop

Humans review and guide AI decisions at critical points.

L

LLM (Large Language Model)

Large language models that understand and generate human language.

M

Machine Learning (ML)

Computers learn from data instead of being manually programmed.

O

On-Premises Hosting

Running AI components on your own infrastructure or data center.

OpenClaw

Open-source framework for agentic AI workflows in enterprise environments.

P

Prompt

The input or instruction you give to an AI model.

R

RAG (Retrieval-Augmented Generation)

AI searches relevant documents first, then generates its response.

T

Temperature (Sampling)

Controls how creative or predictable AI responses are.

Token & Context Window

The unit of measurement for text in AI models and their processing limit.

Tool / Function Calling

AI can invoke external tools and APIs to complete tasks.

Training vs. Inference

Training = AI learns. Inference = AI works.

V

Vector Database

A database optimized for fast similarity search.