AI Stack #

The AI Stack describes the layers required to build an end-to-end AI system, from infrastructure at the bottom to user-facing applications at the top.

Different organisations represent the AI stack differently; this is a simplified conceptual view for learning.

Each layer depends on the one below it.

graph TB

    subgraph APP["Applications"]
        A[User Interfaces & Integrations]
    end

    subgraph ORCH["Orchestration"]
        O[Workflows • Agents • Control Logic]
    end

    subgraph DATA["Data"]
        D[Data Sources • Pipelines • Vector DBs]
    end

    subgraph MODEL["Models"]
        M[ML • DL • Foundation Models • LLMs]
    end

    subgraph INFRA["Infrastructure"]
        I[Cloud • On-prem • GPUs • Storage]
    end

    %% Styling
    style APP fill:#FFCCBC
    style ORCH fill:#90CAF9
    style DATA fill:#BBDEFB
    style MODEL fill:#C8E6C9
    style INFRA fill:#E1F5FE

    style A fill:#FFE0B2
    style O fill:#B3E5FC
    style D fill:#E3F2FD
    style M fill:#DCEDC8
    style I fill:#E1F5FE

1. Infrastructure #

The foundation that provides compute and storage.

• Cloud (AWS, Azure, GCP)
• On-premise servers
• Local machines (laptops, edge devices)
• CPUs, GPUs, TPUs
• Networking and storage

Without infrastructure, AI cannot run or scale.

2. Models #

The intelligence layer of the system.

• Open-source or proprietary models
• Small models or large models (LLMs)
• General-purpose or specialised models
• Examples:
  • Classical ML models
  • Neural Networks
  • Deep Learning models
  • Foundation Models
  • LLMs

Models transform data into predictions or generated content.

3. Data #

The fuel for AI systems.

• Data sources (databases, APIs, files, sensors)
• Data pipelines (ingestion, cleaning, transformation)
• Structured and unstructured data
• Vector databases for embeddings

Better data usually matters more than bigger models.

4. Orchestration #

The control layer that manages AI behaviour.

• Decides when and how models are used
• Combines:
  • Thinking
  • Execution
  • Review and feedback
• Handles workflows, retries, and tool usage

This is where modern AI systems become intelligent systems, not just models.

5. Applications #

The user-facing layer.

• Interfaces:
  • Text
  • Images
  • Audio
  • Numerical data
• Integrations:
  • Inputs (users, systems)
  • Outputs (dashboards, APIs, actions)

This is the only layer users usually see.

One-Line Summary #

• Infrastructure runs everything
• Models provide intelligence
• Data feeds the models
• Orchestration coordinates behaviour
• Applications deliver value to users

flowchart TB
    INFRA[Infrastructure]
    MODEL[Models]
    DATA[Data]
    ORCH[Orchestration]
    APP[Applications]

    INFRA --> MODEL
    MODEL --> DATA
    DATA --> ORCH
    ORCH --> APP

Home | AI Foundation