By the end of this course you will implement your own AI System.

Decision Making

Realtime decision making is one of the key components when creating a solid Game AI.

Learn how a Finite State Machine works and how to use a Hierarchical Finite State machine in your own game project.


Once a decision was made, the AI Agent needs to move from A to B. Here is where pathfinding comes into play.

Learn how to make your own pathfinder and how to use Godot’s Navmesh in a project.


The last piece of the puzzle of a great AI is also in this course!

Learn how to implement range, field of view, raycast sensors and also AI Agent communication.


I’ve really enjoyed the Ultimate Godot Game AI for Beginners course. The pacing and details were spot on for me. The included project is a great working example of how to integrate everything. Even the bonus material was good to know for the beginner. I would gladly recommend this course to anyone interested in taking their project from just a tech demo to a real game with functional opponents.

-- Todd Z

Everything to get you started with ai in godot

The course covers the following topics:

Part 1 (Game AI Concepts)

  • Introduction to Game AI
  • How does the AI perceive the Environment
  • Decision Making – Finite State Machines
  • From A to B – Pathfinding
  • Sensors – Range, Field of View and Raycast

Part 2 (Full example project)

  • Full Project with Assets
  • Videos explaining different systems

Part 3 (Bonus Section)

  • GDScript Introduction
  • Data Structures for Game AI

This course is for anyone interested learning more about how to implement and use AI Systems in Godot Engine. If you are beginner, medium or advanced, you are covered.

Skills required:

  • Basic understanding of the Godot Engine
  • How to create a Node, a Scene, Script




Meet the Agressive AI! Made to seek & kill with a low chance of retreat, this is the perfect killing machine!

It’s main features are a linear projectile, a range sensor combined with raycast and a low chance of retreat.

Learn how to put together an AI Agent that seeks to kill.


Combining attacking with defense is another way to approach a fight. And this is how this AI operates.

Its main features are rocket projectiles, range/raycast sensors and a 50-50% attack/retreat chance.

Learn how to put together a balanced AI Agent that not only seeks the target but also the nearest good cover.


This AI never attacks directly. By having a long range sensor and no raycast, once the enemy is detected a vantage point will be also determined.

The Tactical AI goes to cover, adjusts for missed hits and fires a long-range projectile.

Learn how to put together an AI Agent that uses long-range projectiles, adjusts for errors.

This is a great course with a lot of useful examples about AI in game development
-- Paul S.


You will get not only the examples, but also a full project that features an AI System ready to be used in a game – yes, yours included!

What you will be getting:

  • Relevant examples for each implementation part
  • Full modular project that you can explore and reuse for your game
  • Hierarchical Finite State Machines are used to implement the AI logic
  • Documented Source Code files in GDScript
  • 3 Types of AIs: Aggressive, Defensive and Tactical
  • 3 Projectile Types: Shell, Rocket and Mortar
  • 2 Types of Firing mode: Normal and Burst
  • 2 Patrol modes: Waypoints, Random Movement (interchangeable at runtime)
  • AI Communication System – propagate information between agents
  • Full Player Movement and Firing System
  • Game Assets – Buildings, Tanks and props
  • Everything properly organized in a tidy file structure, prefabs, modular components and more



About me, your AI Coach

Hey, I’m Adrian! A game developer and youtuber.

I started making games in 2005. Back then, there were no game engines like Unity, Unreal, or Godot.

My experience spans more than just development. During high school I put together and led a team of game developers which participated in national gamedev contests, even won some 😀

At the university, I started a gamedev club to teach students how to develop games.

I also released games on steam, together with a team: Desertland2115, Empire of the Fallen Steel, and War of the Wasteland.

Professionally, I worked with Unity, held presentations about various gamedev topics such as pathfinding, networking, steam – even at Nordic Game Conference in Sweden.

My latest project is Redefine Gamedev, the Youtube channel focused on delivering awesome and unique gamedev content. Under REGA I organized GoGodotJam 2x times for an audience of over 10 000 people.

Now I want to share my knowledge with you, to help you reach your goals faster.

Frequently Asked Questions

When does the course start?
It starts now and ends never. You get unlimited access to it.

What if I’m unhappy with the course
I don’t want you to be unhappy with the product! Contact me up until 30 days from the purchase date and I will fully refund you – no questions asked.

Can this AI work for different types of Games?
If you are creating a FPS game, you can easily replace the tank with 3D characters and it will work.

For some game types, this AI might require adjustments to work.

What versions of Godot does it support?
Currently 3.x. The project will be updated once the 4th version of Godot gets officially released.

What prior knowledge should I have?
If you are Beginner, Medium or Advanced and want to learn more about AI in Godot then this Course is for You!

Why HFSM and not BT or PLANNING?
Finite State Machines are one of the easiest ways to implement AI behavior. They do lack a lot of things and this is where Hierarchical Finite State Machines come.

I’ve chosen this system because it’s used in modern games like Doom 2016 and it’s very easy to understand and use while being powerful.