Platform Overview
Relevant source files
The following files were used as context for generating this wiki page:
AtlasCleanAI Enterprise provides an AI-powered ecosystem for managing professional cleaning services. The platform automates booking lifecycles, cleaner scheduling, and operational intelligence for customers, field cleaners, and management teams.
The system integrates a Node.js backend with AI capabilities powered by OpenAI to optimize dispatching, pricing, and service recommendations. It utilizes a monorepo structure containing the core API, infrastructure definitions, and shared Flutter packages for cross-platform mobile applications.
Sources: README.md:3-13, package.json:1-12
System Architecture
The platform follows a containerized, cloud-native architecture. It uses NGINX or an Application Load Balancer for ingress, a decentralized Node.js backend for business logic, and dedicated tiers for persistence and caching.
High-Level Component Flow
The following diagram illustrates the flow from client requests through the infrastructure tiers to the persistence layer.
Infrastructure Tiers
The platform defines three isolated network tiers to ensure security and environment isolation:
- Public Tier: Hosts Load Balancers and NGINX instances.
- Application Tier: Runs containerized Node.js services on ECS Fargate.
- Data Tier: Contains isolated RDS PostgreSQL and ElastiCache Redis instances.
Sources: infrastructure/aws/cdk/lib/atlasclean-stack.ts:28-48, infrastructure/terraform/main.tf:58-94
Core Platform Capabilities
The platform provides a suite of services centered around cleaning operations and AI intelligence.
| Feature | Description |
|---|---|
| Booking Engine | Manages service lifecycle from request to completion. |
| Workforce Management | Handles cleaner onboarding, scheduling, and assignments. |
| Atlas AI Layer | Provides demand prediction, dynamic pricing, and smart scheduling. |
| Payments | Integrates with Stripe for transactional billing and refunds. |
| Messaging | Supports email, SMS, and Firebase push notifications. |
Sources: README.md:15-23, backend/src/app.ts:135-175
Data Model
The persistence layer uses PostgreSQL, managed via Prisma ORM. The schema defines the relationships between users, cleaners, services, and bookings.
Atlas AI Integration
Atlas AI is the intelligence layer of the platform, utilizing OpenAI's GPT models (defaulting to gpt-4o) to provide operational assistance.
AI Functional Flow
When a booking is created, the system invokes AI use cases to calculate prices and suggest cleaners.
AI Configuration Parameters
The AI service requires specific environment variables for operation:
| Variable | Default | Purpose |
|---|---|---|
OPENAI_API_KEY | — | Required for OpenAI API communication. |
OPENAI_MODEL | gpt-4o | The model used for generating responses. |
ATLAS_API_KEY | — | Enables optional API-key authentication for chat. |
Sources: README.md:95-103, src/services/atlasAI.js:16-18, backend/src/config/index.ts:112-115
Deployment and CI/CD
The platform supports two Infrastructure as Code (IaC) providers: AWS CDK for TypeScript-based definitions and Terraform for provider-agnostic workflows.
- CI Pipeline: Validates backend code, Flutter packages, and performs security scans.
- Staging Deployment: Automated deployment to AWS staging environments for validation.
- Production Deployment: Gated deployment using GitHub Environments and AWS OIDC.
Sources: README.md:105-117, infrastructure/aws/cdk/package.json:6-10, infrastructure/terraform/main.tf:1-15
Summary
AtlasCleanAI Enterprise serves as a production-ready foundation for cleaning service marketplaces. It combines a robust containerized backend with AI-driven operational tools, organized within a monorepo that ensures consistency across infrastructure, API, and mobile client applications.