Dim-Jones
AI

Agentic AI: Designing Autonomous AI Agents

Learn to design, develop and deploy autonomous AI agents capable of reasoning, planning and acting.

Duration

2 days (14h)

Level

Intermediate

Format

On-site, Remote

Objectives

  • Understand the principles of agentic AI and how it differs from classic generative AI
  • Design agent architectures with reasoning loops
  • Implement agents using tools (tool use) and data sources
  • Orchestrate multi-agent systems
  • Evaluate and improve agent reliability in production

Prerequisites

  • Basic programming knowledge (Python recommended)
  • Familiarity with basic AI and LLM concepts

Program

1

Day 1

Agentic AI Fundamentals

  • From LLMs to agents: evolution and key concepts
  • Agentic architectures: ReAct, Plan-and-Execute, Reflexion
  • Tool use: enabling agents to interact with the world
  • Memory and context: short-term vs long-term
  • Hands-on: create an autonomous research agent
2

Day 2

Advanced Systems and Production

  • Multi-agent orchestration and task delegation
  • Grounding and RAG for reliable agents
  • Evaluation, monitoring and guardrails
  • Production patterns: retry, fallback, human-in-the-loop
  • Hands-on: multi-agent system for business automation

Who is this for?

This training is designed for developers and technical architects who want to go beyond simple prompt engineering to build truly autonomous AI systems. Concepts are illustrated with modern frameworks and real-world use cases.

Teaching approach

The training alternates between theory and intensive practice. Participants progressively build increasingly sophisticated agents, starting from a simple agent up to a complete multi-agent system capable of automating business processes.

Interested in this training?

Contact us to get a personalized quote or schedule a session.