Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
This article helps organizations build AI workloads on Azure platform-as-a-service (PaaS) solutions. These services support both generative and nongenerative AI workloads with enterprise-grade security and scalability.
Use generative AI architectures and guides
Generative AI architectures create new content and enable conversational experiences through large language models. The architectures provide different complexity levels to match your organization's needs and technical maturity. You must select the appropriate architecture based on your organization's size, compliance requirements, and existing Azure infrastructure. Here's how:
Start with baseline architectures that provide proven design patterns for production workloads. These architectures include security configurations, networking designs, and operational practices that enterprises need for reliable AI deployments. They address common challenges like model governance, cost management, and data protection.
Article Article type Target organization Baseline Azure OpenAI architecture in an Azure landing zone Architecture Enterprise Baseline Azure OpenAI reference architecture Architecture Any Basic Azure OpenAI reference architecture Architecture Startup Apply operational guidance that supports AI development lifecycle management. These guides establish practices for model deployment, monitoring, and continuous improvement across development environments. They ensure consistent quality and reliability as AI applications evolve.
Article Article type Target organization GenAIOps Guide Any Developing RAG solutions Guide Any Proxy Azure OpenAI usage Guide Any Implement Well-Architected design areas that address specific technical concerns for AI workloads. These design areas provide detailed recommendations for application design, data management, and operational excellence that complement the architectural patterns.
Article Article type Target organization Application design Design area Any Application platform Design area Any Training data design Design area Any Grounding data design Design area Any Data platform Design area Any MLOps and GenAIOps Design area Any Operations Design area Any Test and evaluate Design area Any Responsible AI Design area Any
Use nongenerative AI architectures and guides
Nongenerative AI architectures focus on classification, prediction, and analysis tasks without creating new content. These architectures process existing data to extract insights, automate decisions, and enhance business processes. You must choose architectures that align with your specific data processing and analysis requirements.
Here's how:
Select specialized architectures that address specific use cases for data processing and analysis. These architectures demonstrate proven patterns for common business scenarios like document processing, media analysis, and predictive analytics. They provide implementation guidance for integrating AI capabilities into existing business processes.
Article Article type Target organization Document processing architectures Architecture Any Video and image classification architecture Architecture Any Audio processing architecture Architecture Any Predictive analytics architecture Architecture Any Apply operational frameworks that support machine learning model lifecycle management. These guides establish best practices for model training, deployment, and monitoring that ensure consistent performance and reliability in production environments.
Article Article type Target organization Azure Machine Learning Guide Any MLOps Guide Any
Adopt AI standards
The following articles provide best practices for adopting AI workloads using Azure PaaS solutions. These practices ensure your AI workloads meet requirements for security, governance, and operational excellence across key categories.
- Resource selection - Use best practices to choose appropriate AI services and compute resources.
- Networking - Apply best practices to design secure and performant network connectivity.
- Governance - Follow best practices to establish policies and controls for AI resource management.
- Management - Implement best practices for monitoring and operational practices.
- Security - Enforce best practices for applying security controls and compliance requirements.
Azure resources
| Category | Tool | Description |
|---|---|---|
| Platform Services | Azure AI Foundry | Unified platform for building and deploying generative and nongenerative AI applications |
| Platform Services | Azure OpenAI | Access to OpenAI models with enterprise security and compliance |
| Platform Services | Azure Machine Learning | End-to-end machine learning lifecycle management platform |
| Platform Services | Azure AI Services | Prebuilt AI capabilities for vision, speech, language, and decision making |
| Architecture Guidance | Azure Architecture Center AI/ML | Comprehensive collection of AI and machine learning architecture patterns |
| Best Practices | Well-Architected Framework AI Workloads | Design principles and recommendations for AI workloads on Azure |