Interface LLMUtilityAIConfig

Base configuration for any IUtilityAI implementation.

interface LLMUtilityAIConfig {
    utilityId?: string;
    defaultLanguage?: string;
    resourcePath?: string;
    llmProviderManager: AIModelProviderManager;
    defaultModelId?: string;
    defaultProviderId?: string;
    summarizationModelId?: string;
    classificationModelId?: string;
    keywordModelId?: string;
    sentimentModelId?: string;
    languageDetectionModelId?: string;
    jsonFixerModelId?: string;
    semanticSimilarityModelId?: string;
    textNormalizationModelId?: string;
    readabilityEstimationModelId?: string;
}

Hierarchy (view full)

Properties

utilityId? defaultLanguage? resourcePath? llmProviderManager defaultModelId? defaultProviderId? summarizationModelId? classificationModelId? keywordModelId? sentimentModelId? languageDetectionModelId? jsonFixerModelId? semanticSimilarityModelId? textNormalizationModelId? readabilityEstimationModelId?

Properties

Optional utilityId

utilityId?: string

Unique identifier for this specific utility AI service instance.

Optional defaultLanguage

defaultLanguage?: string

Default language for processing if not specified in method options (e.g., 'en', 'es'). BCP-47 format preferred.

Optional resourcePath

resourcePath?: string

Path to a directory containing resource files (e.g., stop word lists, lexicons, trained models for statistical utilities).

llmProviderManager

llmProviderManager: AIModelProviderManager

Optional defaultModelId

defaultModelId?: string

Optional defaultProviderId

defaultProviderId?: string

Optional summarizationModelId

summarizationModelId?: string

Optional classificationModelId

classificationModelId?: string

Optional keywordModelId

keywordModelId?: string

Optional sentimentModelId

sentimentModelId?: string

Optional languageDetectionModelId

languageDetectionModelId?: string

Optional jsonFixerModelId

jsonFixerModelId?: string

The model a direct parseJsonSafe() call repairs with when it names no llmModelIdForFix. The task methods (classifyText, analyzeSentiment and the others) repair a reply on the model and provider the call ran on.

Optional semanticSimilarityModelId

semanticSimilarityModelId?: string

Optional textNormalizationModelId

textNormalizationModelId?: string

The model normalizeText() calls, in any language; defaultModelId unless given.

Optional readabilityEstimationModelId

readabilityEstimationModelId?: string

Settings

Member Visibility

Theme

On This Page

utilityIddefaultLanguageresourcePathllmProviderManagerdefaultModelIddefaultProviderIdsummarizationModelIdclassificationModelIdkeywordModelIdsentimentModelIdlanguageDetectionModelIdjsonFixerModelIdsemanticSimilarityModelIdtextNormalizationModelIdreadabilityEstimationModelId