top of page
aix logo (white).png

PLATFORM ARCHITECTURE

A modular, scalable architecture designed for secure and enterprise AI deployment.

Architecture Overview


AiX is built on a modular, layered architecture designed for secure, scalable, and enterprise-ready AI deployment.

​

It separates user interaction, application services, AI processing, and data management to ensure flexibility and full control over how data and AI workloads are executed.
 

architecture_overview.png

Core Components of AiX

Each component has a clearly defined responsibility, enabling secure operation, controlled data access, and scalable AI processing.

image.png

Web Service (UI)

Provides a secure, browser-based user interface for accessing AiX features, workflows, and system controls.

image.png

Instant Messaging Server

Handles secure ingestion of messages from external channels, normalizing and routing them for processing and storage.

image.png

Console

Controls infrastructure and service management, including node registration, connectivity, and lifecycle configuration.

image.png

Application Server

Acts as a secure gateway for managing data access, enforcing business rules and controlling all database operations.

image.png

Database (PostgreSQL)

Provides persistent, auditable storage for platform data, knowledge, workflows, and system activity.

image.png

Ai Server

Orchestrates AI capabilities including Knowledge Base, Assistants, and Smartflows across the platform.

image.png

Ai Worker

Executes Smartflows and AI tasks, handling distributed processing and scaling based on workload demand.

image.png

Ai Engine

Integrates with AI models, supporting both local GPU-based models and external services for flexible inference.

Deployment & Network Diagram

AiX is designed for secure enterprise deployment with clearly defined network boundaries, controlled data flow, and strict access control across all components.

Network.png

Deployment Requirements

AiX supports flexible deployment across on-premise, cloud, and hybrid environments, with infrastructure sizing based on workload and AI usage.

Minimum Deployment

Best for:
Evaluation, proof-of-concept, small workloads

Deployment Size:
Single machine

Typical resources:

  • CPU: 8–12 cores

  • RAM: 32 GB

  • Storage: ~1 TB SSD

  • GPU (local models): Optional

Scaled Deployment

Best for:

Pilot usage, multi-user environments, moderate AI workloads
 

Deployment Size:

1–2 machines (optional separation of application and AI processing)
 

Typical resources:

  • CPU: 16+ cores  

  • RAM: 64 GB  

  • Storage: 1–2 TB SSD  

  • GPU (local models): Recommended

Enterprise Deployment

Best for:

Production-ready, high concurrency, large-scale AI operations
 

Deployment Size:

Multi-node architecture with dedicated AI processing nodes
 

Typical resources:

  • CPU: Scalable / multi-node  

  • RAM: 64 GB+ per node

  • Storage: Scalable SSD based on data volume  

  • GPU (local models): Dedicated GPU nodes for high-performance inference

bottom of page