Playtesting AI Suite

Project image

Business Goals

  • Automate game testing.

Challenge

  • Manual game testers perform exhaustive playtesting to find gaps bugs or verify a balance. The playtesters are often inaccurate, prone to biased opinions and fatigue, and can’t play below their skill after mastering a game. Manual playtesting becomes an even more costly, lengthy, and complex task as the game develops and becomes famous.
  • The overall goal is to minimize or eliminate the manual playtesting burden in creating balanced complexity levels, characters, maps, missions, etc.

Results

  • With the involvement of the AI testing kit, the human involvement in the testing of one game level was reduced from 3 hours to 7 minutes.
  • AI agents with different game skill levels play in parallel on many game levels.

Implementation Details

  • Self-learning agents were developed and validated for game levels.
  • Scaling and multiple simultaneous runtimes support were implemented via containerization of the AI workspaces and centralized management of the workloads with Kubernetes.
  • Reporting was automated and delivered directly to game developers, testers, and management.
  • The AI agent skill (low, middle, high) is controlled by the deep reinforcement learning model configuration.

Industry

Service

Type

  • Case Study

Keywords

  • Deep Reinforcement Learning
Roadmap
Solution Design
AI Solutions Architest
Data Collection
Data Engineer
Exploratory Data Analysis
Computer Vision Engineer
Model Development
Computer Vision Engineer
Model Performance Evaluation
Computer Vision Engineer
Hyperparameters Tuning
Computer Vision Engineer
Standalone AI Service Design
AI Solutions Architect, MLOps
Standalone AI Service Coding
MLOps
Deployment Infrastructure Design
AI Solutions Architect, MLOps
Deployment Infrastructure Implementation
MLOps
AI Service Deployment
MLOps
AI Service Integration
MLOps
Model Training Automation
MLOps, Computer Vision Engineer
CI/CD
MLOps
Release

Sign up to receive the project description

    Want to talk?

    Michael Gurbych Director,
    Operations and Finance
    Book a meeting
    Roadmap
    Solution Design
    AI Solutions Architest
    Data Collection
    Data Engineer
    Exploratory Data Analysis
    Computer Vision Engineer
    Model Development
    Computer Vision Engineer
    Model Performance Evaluation
    Computer Vision Engineer
    Hyperparameters Tuning
    Computer Vision Engineer
    Standalone AI Service Design
    AI Solutions Architect, MLOps
    Standalone AI Service Coding
    MLOps
    Deployment Infrastructure Design
    AI Solutions Architect, MLOps
    Deployment Infrastructure Implementation
    MLOps
    AI Service Deployment
    MLOps
    AI Service Integration
    MLOps
    Model Training Automation
    MLOps, Computer Vision Engineer
    CI/CD
    MLOps
    Release