Uploaded August 2026 | Updated September 2026, 2 weeks ago
Part of the Modern AI Apps with Azure Cosmos DB Series.
AI coding agents can generate working code quickly, but working code isn't always production-ready code. This session focuses on reviewing agent-generated Azure Cosmos DB code for scalability, RU efficiency, indexing strategy, and long-term maintainability.
Presenters:
Sajeetharan Sinnathurai - Azure Cosmos DB Principal Program Manager - linkedin.com/in/sajeetharan
Jay Gordon - Azure Cosmos DB Senior Program Manager linkedin.com/in/jaygordon0042
• Start with the Agent Kit — Give your coding agent the context it needs to build Cosmos DB apps the right way - https://aka.ms/CosmosDB-agentkit
• Subscribe to Azure Cosmos DB on YouTube - youtube.com/AzureCosmosDB
• Follow Azure Cosmos DB on X - twitter.com/AzureCosmosDB
• Follow Azure Cosmos DB on LinkedIn - linkedin.com/company/azure-cosmos-db
02:52 Welcome & Session Introduction
03:17 Modern AI Apps with Azure Cosmos DB Series Overview
04:10 Repo Walkthrough & Demo Setup
05:30 Why Scaling Decisions Matter Early
06:10 Cosmos DB Scaling Fundamentals: Data Modeling, Partition Keys, RU Efficiency & Indexing
08:24 Creating Cosmos DB with AI Prompts and Entra Authentication
09:32 Seeding the Database with Sample Data
10:00 Building the Ticketing API with GitHub Copilot Agent
10:52 Choosing the Right AI Model and Understanding Costs
12:35 Understanding Application Queries and Access Patterns
15:44 Using OpenAPI for Load Testing Discovery
16:18 Examining Query Patterns and Identifying Performance Issues
17:25 Introducing Azure Cosmos DB Shell
19:47 Exploring Data with the VS Code Cosmos DB Extension
24:26 Generating Natural Language Queries Against Cosmos DB
26:00 Launching the Application and Generating Load
27:27 Using LoadGen to Stress Test the API
30:29 Measuring RU Consumption and Performance
32:09 The Four-Lens Performance Analysis Framework
35:19 Introducing Azure Cosmos DB Agent Kit
36:10 Installing Agent Kit in VS Code
37:59 Lens #1: Evaluating Data Modeling and Partition Key Design
40:14 Lens #2: Why Are These Queries So Expensive?
42:07 Lens #3: Indexing Policy Optimization
45:26 Lens #4: SDK Best Practices and Production Readiness
47:02 Letting Agent Kit Fix the Application
48:43 Baseline Performance Test (Before Optimization)
50:43 Running the Optimized Application (After Agent Kit)
53:13 RU Savings and Performance Improvements
54:03 Key Takeaways: Partition Keys, RU Costs, Indexing & SDK Design
54:42 Azure Cosmos DB Agent Kit Resources
55:25 Final Thoughts on Cosmos DB Shell and Developer Tools
56:20 Next Episode Preview: Building APIs Faster with Agentic Development
57:36 Agent Kit Roadmap, Open Source Contributions & Multi-Agent Support
58:19 Closing Remarks
#azurecosmosdb #aicoding
[eventID:27409]
Part of the Modern AI Apps with Azure Cosmos DB Series.
AI coding agents can generate working code quickly, but working code isn't always production-ready code. This session focuses on reviewing agent-generated Azure Cosmos DB code for scalability, RU efficiency, indexing strategy, and long-term maintainability.
Presenters:
Sajeetharan Sinnathurai - Azure Cosmos DB Principal Program Manager - linkedin.com/in/sajeetharan
Jay Gordon - Azure Cosmos DB Senior Program Manager linkedin.com/in/jaygordon0042
• Start with the Agent Kit — Give your coding agent the context it needs to build Cosmos DB apps the right way - https://aka.ms/CosmosDB-agentkit
• Subscribe to Azure Cosmos DB on YouTube - youtube.com/AzureCosmosDB
• Follow Azure Cosmos DB on X - twitter.com/AzureCosmosDB
• Follow Azure Cosmos DB on LinkedIn - linkedin.com/company/azure-cosmos-db
02:52 Welcome & Session Introduction
03:17 Modern AI Apps with Azure Cosmos DB Series Overview
04:10 Repo Walkthrough & Demo Setup
05:30 Why Scaling Decisions Matter Early
06:10 Cosmos DB Scaling Fundamentals: Data Modeling, Partition Keys, RU Efficiency & Indexing
08:24 Creating Cosmos DB with AI Prompts and Entra Authentication
09:32 Seeding the Database with Sample Data
10:00 Building the Ticketing API with GitHub Copilot Agent
10:52 Choosing the Right AI Model and Understanding Costs
12:35 Understanding Application Queries and Access Patterns
15:44 Using OpenAPI for Load Testing Discovery
16:18 Examining Query Patterns and Identifying Performance Issues
17:25 Introducing Azure Cosmos DB Shell
19:47 Exploring Data with the VS Code Cosmos DB Extension
24:26 Generating Natural Language Queries Against Cosmos DB
26:00 Launching the Application and Generating Load
27:27 Using LoadGen to Stress Test the API
30:29 Measuring RU Consumption and Performance
32:09 The Four-Lens Performance Analysis Framework
35:19 Introducing Azure Cosmos DB Agent Kit
36:10 Installing Agent Kit in VS Code
37:59 Lens #1: Evaluating Data Modeling and Partition Key Design
40:14 Lens #2: Why Are These Queries So Expensive?
42:07 Lens #3: Indexing Policy Optimization
45:26 Lens #4: SDK Best Practices and Production Readiness
47:02 Letting Agent Kit Fix the Application
48:43 Baseline Performance Test (Before Optimization)
50:43 Running the Optimized Application (After Agent Kit)
53:13 RU Savings and Performance Improvements
54:03 Key Takeaways: Partition Keys, RU Costs, Indexing & SDK Design
54:42 Azure Cosmos DB Agent Kit Resources
55:25 Final Thoughts on Cosmos DB Shell and Developer Tools
56:20 Next Episode Preview: Building APIs Faster with Agentic Development
57:36 Agent Kit Roadmap, Open Source Contributions & Multi-Agent Support
58:19 Closing Remarks
#azurecosmosdb #aicoding
[eventID:27409]
