Cosmos DB Depth
Required: Full operational expertise, not just basic CRUD. You should be able to design and optimize partition keys, tailor indexing policies, and tune Request Unit (RU) consumption to ensure sub-10 ms latencies at scale.
Key Skills:
• Partitioning: Design multi-tenant / time-series schemas; mitigate hot spotting by choosing the right partition key
• Index Management: Define custom indexing paths to optimize query patterns and reduce RU costs
• Throughput Tuning: Auto scale vs. manual scaling; cost-optimization strategies around RU provisioning
• Monitoring & Diagnostics: Use Azure Monitor and Cosmos DB metrics to detect and resolve performance bottlenecks
C#/.NET Stack
Target Version: .NET 8 LTS (we’re upgrading from .NET 6/7)
Key Skills:
• ASP.NET Core Web API best practices, middleware, dependency injection, logging
• Entity Framework Core or direct Azure Cosmos/Amazon DynamoDB client usage
• Asynchronous programming (async/await, IAsyncEnumerable<T>)
• Experience with SignalR or gRPC is a plus for real-time scenarios
Azure SDK Integrations
Critical SDKs:
• Azure Messaging Notification Hubs (or the older Microsoft Azure, Notification Hubs) for push notifications
• Azure Cosmos for database operations with .NET
• Azure Identity for unified authentication across managed identities, service principals, and user credentials
Azure Core Services (Critical Expertise Required):
• Service Bus – Reliable message queuing, pub/sub messaging for distributed systems
• Event Grid – Event-based routing to serverless functions and microservices
• Azure Functions – Serverless compute engine to process events, timers, or HTTP triggers
• Azure Notification Hubs – Mobile push notification platform for cross-device messaging
AWS Equivalents (Greenfield Development Familiarity Required):
• Amazon SNS (Simple Notification Service) – Equivalent to Azure Notification Hubs for pub/sub messaging and mobile push notifications
• Amazon SQS (Simple Queue Service) – Equivalent to Azure Service Bus for distributed, reliable queuing
• AWS Lambda – Equivalent to Azure Functions for serverless compute
• Amazon Event Bridge – Equivalent to Azure Event Grid for event-driven integration across services
Key Expectations:
• Understand both the architecture and the operational nuances of these services in Azure and their AWS counterparts
• Design migration plans where existing Azure-native services (like Event Grid triggers) are being re-architected in AWS (using Event Bridge, SNS/SQS, and Lambda)
API Integration
Current Stack:
• REST behind Azure API Management (APIM) for request routing, security (OAuth 2.0/JWT), and rate-limiting
Future Stack:
• Kong API Gateway: Expect to configure service definitions, plugins for authentication (Key-Auth, OAuth), and request transformations. You should know how to migrate policies from APIM to Kong equivalents
Expectations:
• Define and document OpenAPI/Swagger specs
• Implement client-friendly versioning, CORS, and health-check endpoints
IaC Depth
Mandatory:
• Bicep or ARM templates for core resource deployments (Notification Hubs, Cosmos DB, APIM/Kong infra)
Highly Desirable:
• Terraform (either in Azure or AWS) to manage multi-cloud or hybrid scenarios with modular design, state management, and CI/CD pipelines
Skills:
• Parameterization, modules, secure storage of secrets
• Integration with Azure DevOps, GitHub Actions, or AWS Code Pipeline
Any Graduate