AI Factory — MLOps Landing Zone

A dedicated AI Factory Landing Zone for the full ML lifecycle: data engineering, feature engineering, model training, registry, deployment pipelines, and monitoring — all within a purpose-built subscription.

~$6,120/mo
12
Azure Services
~$6,120
Monthly Cost
99.95%
Databricks SLA
Full MLOps
End-to-End

Overview

The AI Factory pattern creates a purpose-built landing zone for the complete ML lifecycle. Azure Databricks handles data engineering and distributed training, Azure Machine Learning manages experiments, model registry, and pipelines, while Azure OpenAI provides LLM inference endpoints.

This is the go-to pattern for enterprise ML teams that need CI/CD for models, feature engineering at scale, and automated retraining pipelines.

Key Characteristics
  • Databricks Premium — VNet-injected, Spark-based training and ETL
  • AML Workspace — Pipeline orchestration, experiment tracking, model registry
  • Dual ADLS Accounts — Separate feature store and model storage
  • ACR Premium — Container registry for ML model images
  • 120-day Log Retention — Audit compliance for ML experiments
  • CI/CD — GitHub Actions / Azure DevOps pipeline automation
Databricks Config
TierPremium
VNet InjectionEnabled
Public Subnet/22 delegated
Private Subnet/22 delegated
Feature Store Containers
feature-store training-data model-artifacts experiment-outputs
Model Storage Containers
registered-models model-packages serving-artifacts

Architecture Diagram

graph TB
    subgraph Factory["AI Factory VNet 10.20.0.0/16"]
        subgraph DBX["Databricks VNet-Injected"]
            DBPUB["Public Subnet /22"]
            DBPRV["Private Subnet /22"]
        end
        subgraph Compute["ML Compute"]
            AML["Azure ML Workspace<br/>Pipelines & Registry"]
            MLCOMP["ML Compute /22"]
        end
        subgraph Data["Data Layer"]
            FEAT["ADLS Gen2<br/>Feature Store"]
            MODELS["ADLS Gen2<br/>Model Storage"]
        end
        subgraph AI["AI Services"]
            AOAI["Azure OpenAI<br/>LLM Inference"]
            ACR["Container Registry<br/>Premium"]
        end
        subgraph Sec["Security & Monitoring"]
            KV["Key Vault"]
            MON["Log Analytics<br/>120-day retention"]
            AI2["App Insights"]
        end
    end
    subgraph CICD["CI/CD"]
        GH["GitHub Actions /<br/>Azure DevOps"]
    end
    DBX -->|"ETL / Training"| FEAT
    DBX -->|"Feature Eng"| MODELS
    AML --> FEAT
    AML --> MODELS
    AML --> ACR
    AML --> KV
    AML --> AOAI
    GH -->|"Pipeline Trigger"| AML
    AOAI --> MON
    AML --> MON
    DBX --> MON
          

Bill of Materials

#ServiceResource NameSKU / TierPurposeMonthly Cost
1Azure Machine Learning{base}-{env}-amlWorkspace + ACRPipelines, experiments, model registry$0*
2Azure Databricks{base}-{env}-dbwPremium, VNet-injectedSpark training, ETL, feature engineering~$2,500
3Azure OpenAI{base}-{env}-openaiS0LLM inference endpoints~$1,800
4ADLS Gen2 (Features){base}{env}featuresStandard LRS, HNSFeature store, training data~$60
5ADLS Gen2 (Models){base}{env}modelsStandard LRS, HNSRegistered models, packages, serving artifacts~$40
6Container Registry{base}{env}acrPremium, 30-day retentionML model container images$167
7Key Vault{base}-{env}-kvStandardKeys, connection strings~$5
8Log Analytics{base}-{env}-lawPerGB2018, 120-dayAudit logging, experiment tracking~$35
9Application Insights{base}-{env}-aiWorkspace-basedPipeline and model telemetryIncl.
10VNet{base}-{env}-vnet/16 with 5 subnetsNetwork backbone for factoryFree
11Private DNS Zones6 zonesGlobalPrivate name resolution~$3
12Private Endpoints4 endpointsOpenAI, Features-Blob, KV, ACR~$29
Estimated Total (Moderate Production)~$6,120/mo

* AML workspace free; compute instances billed separately. Databricks cost based on Premium DBU consumption.

Service Breakdown

Azure Databricks
Premium~$2,500/mo

VNet-injected Databricks workspace for Spark-based distributed training, ETL pipelines, and feature engineering. Public and private subnets each /22. No public Databricks infrastructure exposed.

Azure Machine Learning
WorkspaceFree*

Full MLOps workspace linked to ACR, ADLS, Key Vault, and App Insights. Pipeline orchestration, experiment tracking, model registry with versioning and staged deployments.

Azure OpenAI
S0~$1,800/mo

LLM inference endpoints for the AI Factory. Used for generating embeddings, fine-tuning evaluation, and production inference. Private endpoint access only.

Dual ADLS Gen2
Standard LRS~$100/mo

Two separate accounts: (1) Feature Store for computed features, training data, and experiment outputs; (2) Model Storage for registered models, packages, and serving artifacts.

Container Registry
Premium$167/mo

Premium ACR for ML model container images. Admin user disabled (managed identity only). 30-day image retention policy. Geo-replicated for HA.

Monitoring (120-day)
PerGB2018~$35/mo

Log Analytics with 120-day retention for audit compliance. Tracks all Databricks jobs, AML experiments, model deployments, and OpenAI usage.

Security & Networking

Network
  • Databricks VNet-injected — no public infrastructure
  • All data services behind Private Endpoints
  • 6 Private DNS Zones for full private resolution
  • /16 VNet with 5 subnets
  • 120-day log retention for audit
Identity
  • Container Registry admin user disabled (managed identity only)
  • Key Vault with RBAC authorization + purge protection
  • AML workspace linked to ACR, ADLS, KV, App Insights
  • Managed Identities for all service auth
  • CI/CD via GitHub Actions / Azure DevOps

Use Cases

Large-Scale ML Pipelines

Databricks for distributed training, AML for orchestration. Handles petabyte-scale data processing and model training.

Continuous Retraining

Automated pipelines triggered by data drift detection. Retrain, validate, and deploy models without manual intervention.

Enterprise AI Product Teams

Full MLOps lifecycle in one landing zone. From data engineering to production model serving.

Feature Engineering at Scale

Spark-based feature computation on Databricks with ADLS Gen2 feature store. Delta Lake for ACID transactions.

Constraints & Considerations

ConstraintMitigation
Databricks Premium cost is significantUse job clusters (auto-terminate), spot instances for training
VNet-injected Databricks consumes large IP rangesPlan /22 subnets; ensure VNet has capacity
Complex to operate (many moving parts)Use AML pipelines for orchestration; GitOps for config
ACR Premium is expensive for low usageConsider Standard tier if geo-replication not needed
Ready to Deploy This Pattern?

Generate a complete deployment spec sheet with GitHub Actions workflow, Bicep file structure, and prerequisite checklist.

Generate Deployment Spec Sheet