If you have been testing chatbots and LLM prompts and asking how to turn them into real results, this is the next step. Our AWS-authorized Agentic AI Foundations course provides agentic AI training that goes beyond conversation and into measurable impact.
Agentic AI Foundations (AWS Classroom Training) turns prompts into agents: systems that plan, take actions, coordinate tools, and close the loop with measurable outcomes. Now available, this one-day course shows you how to design and evaluate agentic AI architectures using services like Amazon Q, Kiro, Amazon Bedrock Agents, and Amazon Bedrock AgentCore. You will leave with a mental model, a practical toolkit, and hands-on experience building a small but mighty agent you can extend at work.
Why Agentic AI, and Why Now?
LLMs changed what software can understand. Agents change what software can accomplish. They combine reasoning, tool use, memory, and multi-step workflows to execute objectives, including triaging tickets, enriching data, and orchestrating cloud operations.
Agentic AI Foundations helps you go from “chat” to results:
- Blueprints, not buzzwords.
Learn core patterns (workflow, autonomous, hybrid), when each fits, and how they map to AWS services. - Vendor-backed fundamentals.
Understand Amazon Bedrock Agents and AgentCore, what they do, where they shine, and how to implement safely. - Dev velocity.
See where Amazon Q Developer, Amazon Q Business, and Kiro accelerate design, coding, and integration.
Curious how Agentic AI differs from chatbots? Check out the course here →
What You’ll Build and Do (Hands-On)
This is not a passive overview. You will practice designing an agent, defining goals, tools, plans, and guardrails. You will implement a prototype with Amazon Bedrock Agents and AgentCore, connect knowledge with Amazon Bedrock Knowledge Bases, enforce safety with Amazon Bedrock Guardrails, and observe behavior through traces and logs so you can debug, improve, and ship.
Outcome: a working agent scaffold you can adapt for internal workflows, such as support triage, content assembly, data curation, or DevOps runbooks.
Who This Agentic AI Course Is For
- Software developers who are new to agentic AI and want practical foundations
- Technical professionals exploring agent capabilities and core components
- Development teams evaluating agent types and AWS service choices
- AWS users expanding into agents, including Amazon Q, Bedrock Agents, and AgentCore
Prerequisites:
- Generative AI Essentials (or equivalent experience)
- Basic AWS knowledge and software development experience
Program Details at a Glance:
- Level: Fundamental
- Format: AWS-authorized classroom training (instructor-led virtual available)
- Duration: 1 day
- Activities: Presentations, hands-on lab, and group exercises
Deep Dive: What You’ll Learn in Agentic AI Foundations
1) From LLMs to Agents
You will ground yourself in the strengths and limits of large language models, how they reason, where they hallucinate, and why context and tools matter. You will then trace the innovations that enable agents, including tool use, planning loops, memory, and retrieval. By the end of this section, the evolution from single-turn chat to goal-seeking systems will feel clear and inevitable, and you will have a shared vocabulary to discuss agent capabilities with both engineers and stakeholders.
2) Exploring Agentic AI
You will define what makes a system truly agentic, including autonomy, tool orchestration, and iterative reasoning, and you will survey the major agent types. You will examine where each shines across development, operations, customer experience, and knowledge work, and you will translate abstract ideas into concrete scenarios such as automated ticket triage, data enrichment pipelines, and cloud remediation workflows.
3) Agentic Workflows
You will explore workflow patterns that emphasize determinism, auditability, and approvals, and you will see how to orchestrate multi-step tasks without handing over full autonomy. An overview of Amazon Bedrock flows helps you recognize when a workflow-first approach is the right choice and when to complement it with autonomous subroutines for flexibility, which helps you balance safety and speed.
4) Autonomous Agents
This module explains plan, act, and observe cycles and shows how strategies like ReAct and ReWOO structure complex reasoning. You will look at multi-agent collaboration patterns where specialized agents coordinate to cover retrieval, planning, execution, and review. You will also learn where autonomy should be constrained with policies, approvals, and guardrails so exploration never compromises compliance or reliability.
5) Amazon Q and Development Tooling
You will see how Amazon Q Developer accelerates coding, integration scaffolds, and test generation, while Amazon Q Business brings enterprise search and retrieval to your agent workflows. You will also work with Kiro, an AI-powered IDE for spec-driven development, to model goals, tools, contracts, and tests so your agents start with clarity and quality gates rather than ad hoc prompts.
6) Agentic AI with Amazon Bedrock
You will get a practical understanding of Amazon Bedrock Agents, including capabilities, invocation patterns, and common architectures, alongside AgentCore, which provides the primitives for tool use and orchestration. In the hands-on lab, you will connect an agent to Knowledge Bases for context and add Guardrails for safety, producing a portable scaffold you can iterate on after class.
7) DIY Solutions in Production
This section focuses on shipping. You will examine implementation patterns and reference architectures that connect agents to real systems, then dive into observability and monitoring to capture traces, metrics, and evaluations that make iterative improvement possible. You will also learn interoperability techniques for composing agents with services and workflows so teams can scale capability without reinventing the wheel.
8) Wrap-Up
You will leave with curated next steps, a checklist for evaluating your first production use case, and additional resources aligned to the AWS GenAI stack. The goal is confidence. You will know which pattern to pick, which services to reach for, and how to move from proof of concept to dependable delivery.
Choosing the Right Pattern: Workflow vs. Autonomous vs. Hybrid
Workflow agents
- Best for: compliance, financial ops, repeatable processes
- Why: predictable paths, strong auditability, easier approvals
Autonomous agents
- Best for: exploratory tasks, data collection/enrichment, prototyping
- Why: flexible planning, tool selection, multi-step iteration
Hybrid agents
- Best for: complex enterprise use cases
- Why: combine workflow guardrails with autonomous sub-tasks
You’ll practice evaluating your use case and choosing the minimal-complexity option that still delivers outcomes.
Career-Boosting AI Skills You’ll Walk Away With
You will leave speaking the language of agentic systems with both stakeholders and leadership, and you will be able to compare and justify AWS service choices for agentic architectures with confidence. You will prototype quickly using Amazon Q and Kiro, then harden those solutions with Bedrock Agents and AgentCore. From day one, you will design with observability, safety, and interoperability in mind, so your agents are not just clever but also reliable and ready for production.
Great prep for roles like:
- Software Developer (AI and Agents)
- Cloud AI Developer
- Platform or Automation Engineer (GenAI)
- Solutions Engineer or Architect (AI)
How This Fits Your AI Learning Path
Round out your skill set with LearnQuest’s broader AWS + AI catalog:
- Generative AI Essentials on AWS – Core GenAI concepts and services
- AWS Cloud Practitioner Essentials – Cloud fundamentals and certification prep
- AWS Certified AI Practitioner – Validate your AI/ML/GenAI knowledge
Explore our AI Training page for more offerings across AI fundamentals, data science, and Cloud Machine Learning.
Get Started with Agentic AI Fundamentals Today
Ready to design agents that plan, call tools, and deliver real outcomes on AWS?
Frequently Asked Questions
What is agentic AI?
Agentic AI is a design approach where AI systems pursue goals by planning tasks, calling tools or APIs, observing the results, and iterating until the objective is met. Instead of only responding with text, an agent can take actions that change systems or data.
Is this beginner-friendly?
Yes, “Fundamental” level for practitioners with GenAI basics and AWS familiarity.
How technical is the lab?
You’ll follow a guided build using Bedrock Agents/AgentCore with Knowledge Bases and Guardrails, enough to understand the moving parts and extend them later.
Will I learn when not to use autonomous agents?
Absolutely. You’ll compare workflow vs. autonomous vs. hybrid, with selection criteria and risk controls.
Can my team take this together?
Yes. We offer private deliveries tailored to your environment and goals.
What should I review beforehand?
Brush up on Generative AI Essentials (or equivalent), AWS fundamentals, and basic software development.