High-Level Architecture
Relevant source files
The following files were used as context for generating this wiki page:
Aether-X Ultimate is a modular, microservices-based AI platform designed for content generation, no-code application development, and secure data operations. The architecture is built to be "unrestricted," integrating advanced AI models with a robust security layer and a scalable infrastructure stack that supports both SQL and NoSQL data persistence.
The system is containerized using Docker and orchestrated via Kubernetes for production environments. It leverages a centralized API Gateway to handle request routing, rate limiting, and service discovery across eight core microservices.
Sources: README.md:3-10, IMPLEMENTATION_COMPLETE.md:5-15
Microservices Ecosystem
The platform consists of specialized microservices that communicate over a shared network. Each service is independently deployable and responsible for a specific domain of the platform's functionality.
Core Services and Port Mapping
| Service | Port | Primary Responsibility |
|---|---|---|
| API Gateway | 80 | Request routing, load balancing, and rate limiting |
| Text Generation | 8000 | GPT-2 based text synthesis and batch processing |
| Image Generation | 8001 | Stable Diffusion-based image creation |
| Video Generation | 8002 | Video synthesis with Text-to-Speech (TTS) |
| AI Models | 8003 | Specialized domain models and LLM management |
| No-Code Dev | 8004 | Natural language to app/workflow conversion |
| Security | 8005 | JWT auth, AES encryption, and security protocols |
| Data Processing | 8006 | ETL pipelines and Kafka-based streaming |
Sources: README.md:13-21, IMPLEMENTATION_COMPLETE.md:18-26
Request Flow Architecture
The following diagram illustrates the path of a user request through the API Gateway to the various backend microservices and their respective dependencies.
This diagram shows the routing logic from the external entry point to internal services and their persistent storage backends.
Sources: docker-compose.yml:5-150, services/api-gateway/app.py:15-24
System Infrastructure
The infrastructure layer provides the backbone for data persistence, message queuing, and system observability. It is designed for high availability and low latency through the use of specialized database engines for different data types.
Persistence Layer
The platform employs a multi-database strategy to optimize for specific workloads:
- Relational (SQL): MySQL is used for user management, while PostgreSQL stores complex workflow definitions and relational app data.
- Document/NoSQL: MongoDB handles unstructured content storage and metadata.
- Columnar: Cassandra provides high-availability storage for large-scale data.
- Search & Cache: Elasticsearch powers full-text search and analytics, while Redis provides session storage and caching.
Sources: README.md:73-81, infrastructure/databases/init-databases.sh:5-150
Messaging and Streaming
Data streaming and asynchronous processing are handled by Apache Kafka. The infrastructure includes a Schema Registry to manage data evolution and Kafka Connect for integrating external data sources.
Sources: infrastructure/kafka/docker-compose.yml:5-50, infrastructure/databases/docker-compose.yml:106-130
API Gateway & Security Logic
The API Gateway acts as the single entry point for all client communication. It implements a proxy mechanism that forwards requests to the appropriate internal service while applying cross-cutting concerns.
Gateway Responsibilities
- Rate Limiting: Implements request throttling based on client IP to prevent service abuse.
- Logging: Captures request methods, paths, status codes, and durations for audit trails.
- Proxying: Transparently maps public
/api/v1/routes to internal microservice endpoints.
Sources: services/api-gateway/app.py:27-100
The sequence diagram highlights the authentication flow followed by a proxied request to a content generation service.
Sources: services/api-gateway/app.py:145-160, IMPLEMENTATION_COMPLETE.md:52-60
Deployment & Scalability
The system is designed for horizontal scalability using Kubernetes. Deployments are defined with resource limits and liveness/readiness probes to ensure self-healing.
Kubernetes Resource Management
| Resource | Replicas | CPU Request/Limit | Memory Request/Limit |
|---|---|---|---|
| Text Gen | 3 | 250m / 500m | 512Mi / 1Gi |
| Image Gen | 3 | 500m / 1000m | 1Gi / 2Gi |
| Video Gen | 2 | 500m / 1000m | 1Gi / 2Gi |
| AI Models | 2 | 500m / 1000m | 1Gi / 2Gi |
Sources: k8s/deployment.yaml:1-120
Monitoring Stack
The platform includes a built-in monitoring suite for observability:
- Prometheus: Scrapes metrics from services to monitor health and performance.
- Grafana: Provides visual dashboards for real-time monitoring.
- Kibana: Used for log visualization and searching through data indexed in Elasticsearch.
Sources: docker-compose.yml:153-185, IMPLEMENTATION_COMPLETE.md:76-85
Summary
The High-Level Architecture of Aether-X Ultimate represents a decoupled, polyglot persistence system. By separating content generation, no-code logic, and security into distinct microservices, the platform achieves high modularity. The integration of a centralized API Gateway and a comprehensive monitoring stack ensures that the system remains observable and manageable even as it scales across a Kubernetes cluster.