Dynamic AI Pricing

Relevant source files

The following files were used as context for generating this wiki page:

• backend/src/application/use-cases/ai/DynamicPricingUseCase.ts
• backend/src/infrastructure/services/AtlasAIService.ts
• backend/src/interfaces/controllers/AIController.ts
• backend/src/application/use-cases/booking/CreateBookingUseCase.ts
• backend/src/infrastructure/database/prisma/schema.prisma
• backend/src/app.ts

Dynamic AI Pricing calculates service costs based on real-time factors including service type, duration, location, and demand. The system integrates with the AtlasAIService to provide optimized pricing for bookings, ensuring rates reflect current market conditions and resource availability.

The system primary serves the CreateBookingUseCase by generating accurate price estimates before a booking is finalized. It also provides a public API endpoint through the AIController for real-time pricing queries.

Sources: backend/src/application/use-cases/ai/DynamicPricingUseCase.ts, backend/src/application/use-cases/booking/CreateBookingUseCase.ts:33-40

Architecture and Data Flow

Dynamic pricing logic resides within the DynamicPricingUseCase. It coordinates data between the application layer and the AI infrastructure service.

Execution Flow

1. The AIController or CreateBookingUseCase invokes the pricing logic.
2. The DynamicPricingUseCase calculates initial base rates and duration-based costs.
3. The system sends these variables to the AtlasAIService.
4. The AtlasAIService attempts to contact an external AI API for optimized pricing.
5. If the external service is unavailable, the system executes local heuristic fallbacks.

The diagram shows the request path from external interfaces through the use case to the AI service layer.
Sources: backend/src/application/use-cases/ai/DynamicPricingUseCase.ts:16-36, backend/src/infrastructure/services/AtlasAIService.ts:109-138

Core Components

DynamicPricingUseCase

This class calculates the preliminary costs and applies initial multipliers. It maps service types to specific base rates and identifies weekend surges.

Service Type	Base Rate (USD)
standard	80
deep	140
moveout	200
commercial	250
default	100

Sources: backend/src/application/use-cases/ai/DynamicPricingUseCase.ts:18-24

Pricing Multipliers

The system applies several factors to the base cost:

• Duration: Calculated as hours = minutes / 60, with a minimum of 1 hour.
• Extras: Each extra service item adds 15 USD.
• Weekend: A 1.15x multiplier applies if the start date falls on a weekend.
• Rush Hour: A 1.12x multiplier applies for starts during 07:00-08:59 or 17:00-19:59.
• Commercial Boost: Commercial services receive a 1.25x multiplier in fallback mode.

Sources: backend/src/application/use-cases/ai/DynamicPricingUseCase.ts:25-28, backend/src/infrastructure/services/AtlasAIService.ts:122-123

API Implementation

The pricing system is exposed via the /api/v1/ai/pricing/dynamic endpoint.

Pricing Request Structure

interface CalculateDynamicPriceInput {
  baseRate: number;
  hours: number;
  extrasCost: number;
  serviceType: string;
  postalCode: string;
  scheduledStart: Date;
  demandMultiplier: number;
}

Sources: backend/src/infrastructure/services/AtlasAIService.ts:32-40

Controller Integration

The AIController validates required fields before executing the pricing logic.

  dynamicPricing = async (request: Request, response: Response, next: NextFunction): Promise<void> => {
    try {
      const { address, serviceType, scheduledStart, scheduledEnd, extras = [] } = request.body;
      if (!address || !serviceType || !scheduledStart || !scheduledEnd) {
        throw new ValidationError('address, serviceType, scheduledStart, and scheduledEnd are required');
      }

      const pricing = await this.dynamicPricingUseCase.execute({
        address,
        serviceType,
        scheduledStart: new Date(scheduledStart),
        scheduledEnd: new Date(scheduledEnd),
        extras,
      });
      response.status(200).json({ success: true, data: pricing });
    } catch (error) {
      next(error);
    }
  };

Sources: backend/src/interfaces/controllers/AIController.ts:31-50

Data Persistence

Pricing results and factors are stored within the Booking model in the aiMetadata JSON field. This allows for historical auditing of how a specific price was reached.

Field	Type	Description
amount	Decimal(10,2)	Final calculated price.
currency	String	Currency code (defaults to USD).
aiMetadata	Json	Stores pricingFactors such as demandMultiplier and rushHour.

Sources: backend/src/infrastructure/database/prisma/schema.prisma:103-105, backend/src/infrastructure/services/AtlasAIService.ts:130-137

Summary

Dynamic AI Pricing provides a flexible billing mechanism that adjusts to service complexity and demand. By combining fixed base rates with dynamic AI-driven multipliers and local fallbacks, the system maintains consistent pricing operations even during external service interruptions.