Structuring Domain Interactions: Implementing Data Transfer Objects in NestJS

Standardizing Business Logic

In our crm-saas-backend project, we often find ourselves managing various customer touchpoints. As the system grows, passing raw objects or loose data structures between services leads to "hidden" dependencies and runtime errors that are difficult to debug. To solve this, we recently implemented a formalized Data Transfer Object (DTO) pattern to strictly define our interaction types.

The Problem: Loose Data Structures

Previously, tracking interactions like calls, meetings, or messages relied on inferred data shapes. This caused several issues:

1. Type ambiguity: Passing incomplete data objects to processing services.
2. Validation drift: Different parts of the app checked for different properties.
3. Documentation gaps: Swagger integration remained manual and often outdated.

The Solution: Strongly Typed DTOs

By leveraging TypeScript classes and NestJS decorators, we created a clear contract for all incoming interaction data. This ensures that any interaction (call, meeting, or message) conforms to an expected schema before it even reaches the business logic layer.

import { IsEnum, IsString, IsDateString } from 'class-validator';
import { ApiProperty } from '@nestjs/swagger';

enum InteractionType {
  CALL = 'call',
  MEETING = 'meeting',
  MESSAGE = 'message',
}

export class CreateInteractionDto {
  @ApiProperty({ enum: InteractionType })
  @IsEnum(InteractionType)
  type: InteractionType;

  @ApiProperty()
  @IsString()
  summary: string;

  @ApiProperty()
  @IsDateString()
  scheduledAt: string;
}

Benefits for the Team

1. Self-Documenting APIs: Since we use Swagger, the ApiProperty decorators automatically generate accurate API documentation. No more manual updates.
2. Early Validation: The NestJS ValidationPipe intercepts requests at the controller level, rejecting invalid types (e.g., an unsupported interaction category) before execution.
3. Developer Experience: TypeScript intellisense now correctly guides developers when they create new interactions, reducing the likelihood of typos.

Key Insight

Think of DTOs as the "receptionist" for your application logic. Instead of allowing raw data to wander through your codebase, you place a receptionist at the door who checks that everyone has the right credentials. By centralizing these definitions, you ensure that your core business logic receives only the clean, validated data it expects.

Next Steps

1. Audit your current endpoints to identify where plain objects are being passed.
2. Define a base DTO class for common fields to reduce duplication.
3. Run a build to verify that your Swagger documentation reflects these new structures automatically.

Generated with Gitvlg.com