Julian Gomez

A previous attempt to turn VORSPRUNGatwork's consulting approach into an experiential simulation game failed despite five-figure investment. The team was skeptical about game-based learning. I got the project without knowing this history.

Iterative development with continuous stakeholder feedback. Role-switching mechanics combined with VUCA event simulation. Players directly experience how personal mindset and team dynamics work under pressure.

Successfully deployed in corporate training sessions. From skeptical team to working product. Creative perspective shifts solve problems where conventional approaches fail.

• Personality as a basis for decisions: Behavior Trees and State Machines reliably control agent behavior at the action level. Why one character decides differently than another in the same situation is not represented there. Advancing the decision-making behavior of agents (Player/NPCs) and bringing it closer to human decision patterns is the central design challenge.
• Technical feasibility: A more realistic personality model processes many traits across multiple stages. Implementing this under the real-time conditions of a game, with limited computing power and a fixed frame budget, determines whether the approach is practically viable or remains theoretical.
• Comprehensibility and integration: Personality in games involves many factors simultaneously and requires connections to different components of a game engine. Building a system with this scope that remains comprehensible and easy to integrate is a design challenge in its own right.

• Personality profile: Each character receives a profile during development from four psychological categories: emotional, social, cognitive, and moral. The system derives its decisions from this profile.
• Autonomous interpretation: The goal is that new situations don't require new scripts, because the character interprets them on its own.

• Transparency: Every decision outputs a full score breakdown. Developers can debug and predict character behavior.
• Performance: The LOD system scales evaluation depth by relevance across three tiers (Full / Reduced / Minimal) and targets 200 concurrent agents at 60 fps.
• No-Code Configuration: The core product is designed for code-free setup. Characters, context, and connections can be configured via the Unity Inspector without requiring C# knowledge.
• Extensibility: For custom requirements, direct access to the decision logic is available. Custom evaluators, trait definitions, and algorithms can be integrated without modifying the core framework.