← Back to selected work

AI & GAMEPLAY PROGRAMMING

The NegativeEdge

A fast-paced third-person arena shooter where I built the enemy AI architecture, enemy archetypes, wave system and encounter-control systems in C++.

Unreal Engine 5C++Enemy AIEQSWave SystemArena Combat
FAST-PACED ARENA COMBAT
RoleAI & Gameplay Programmer
EngineUnreal Engine 5
LanguageC++
FocusEnemy AI • Encounters

The Negative Edge is built around speed, mobility and surviving escalating waves of enemies in stylized black-and-white arenas.

The player uses advanced movement such as grappling, wall running and rail grinding, so the AI had to remain readable and threatening without creating unfair attack overlap.

Arena combat and movement
Arena combat and movement

I created the enemy foundation and specialized it into four distinct behaviors.

01

Melee

Close-range sword enemies that aggressively pursue and pressure the player.

02

Ranged

Distance-focused enemies that reposition and attack from safer ranges.

03

Explosive

Rush-down units that charge the player and detonate, forcing immediate reactions.

04

Agile Flying

Highly mobile aerial units that can activate totems which temporarily protect other enemies.

Enemy archetypes
Enemy archetypes
WAVES

Wave System

Controls enemy composition, spawning and progression to escalate encounter difficulty.

FAIRNESS

Attack scheduling

Limits how many enemies can commit to attacks at the same time so pressure stays high without becoming unreadable.

POSITIONING

Custom EQS backbone

Coordinates movement points, attack positions and contextual decisions across enemy types.

ARENAS

Zone & area management

Splits encounter spaces into meaningful regions used by spawning and AI positioning logic.

Arena regions and AI area management
Arena regions and AI area management

The core goal was to avoid four completely separate enemy implementations.

  • Modular base classes: Shared enemy/controller responsibilities provide common behavior without locking every archetype to the same strategy.
  • System integration: Wave pacing, EQS positioning and attack scheduling work together instead of competing for control.
  • Iteration: New enemy behavior can be added while preserving the same encounter-level rules.
  • Performance awareness: Zone-based logic helps keep AI decisions focused on the parts of the arena that actually matter.