Enhance company service with AI recommendation caching and onboarding improvements

- Updated the `companyService` to include Redis caching for AI recommendations, improving performance and reducing redundant AI calls.
- Introduced asynchronous AI onboarding triggered after company profile updates, enhancing user experience by offloading processing.
- Added configuration for recommendation cache TTL in the `AISummarizerConfig`, allowing for flexible cache management.
- Implemented methods for caching, retrieving, and invalidating AI recommendations in the `companyService`, ensuring efficient data handling.

This update enhances the company's AI recommendation capabilities, providing faster responses and a more efficient onboarding process.

cmd/web/bootstrap/config.go

@@ -69,6 +69,8 @@ type AISummarizerConfig struct {
 	APIRetryCount   int           `env:"AI_SUMMARIZER_API_RETRY_COUNT" envDefault:"2"`
 	APIRetryDelay   time.Duration `env:"AI_SUMMARIZER_API_RETRY_DELAY" envDefault:"3s"`
 	DefaultLanguage string        `env:"AI_SUMMARIZER_DEFAULT_LANGUAGE" envDefault:"en"`
+	// RecommendationCacheTTL caches AI /recommend responses per company. Set to 0 to disable.
+	RecommendationCacheTTL time.Duration `env:"AI_RECOMMENDATION_CACHE_TTL" envDefault:"15m"`
 
 	// MinIO storage settings
 	MinioEndpoint  string `env:"AI_SUMMARIZER_MINIO_ENDPOINT" envDefault:""`