src/system_tests/profiling/config/providers_comparison.yaml

# Provider Comparison Experiment
# Compare different LLM providers (OpenAI, Azure, WatsonX) with the same mode

profiling:
  # Single mode to test across all providers
  modes:
    - "balanced"
  
  tasks:
    - "test_get_top_account_by_revenue_stream"
  
  runs: 3
  
  langfuse:
    host: "https://cloud.langfuse.com"
    retry:
      max_attempts: 10
      initial_delay: 2.0

experiment:
  name: "providers_comparison"
  description: "Compare OpenAI vs Azure using balanced mode"
  
  runs:    
    - name: "azure_balanced"
      test_id: "settings.azure.toml:balanced:test_get_top_account_by_revenue_stream"
      iterations: 3
      output: "experiments/azure_balanced_{{timestamp}}.json"
      env:
        MODEL_NAME: null
    
  comparison:
    generate_html: true
    html_output: "experiments/comparison.html"
    auto_open: false