5:00 PM - 6:30 PM (UTC)

Join us to learn how to effectively deploy an intelligent microservice-based application into Azure Kubernetes Service (AKS) and integrate with Azure OpenAI. We’ll walk through setting up the application environment, deploying Azure OpenAI services and configuring a deployment of an existing model, and integrating that with our application.

5:00 PM - 6:30 PM (UTC)

Join us to learn how to run open-source Large Language Models (LLMs) with HTTP-based inference endpoints inside your AKS cluster using the Kubernetes AI Toolchain Operator (KAITO). We’ll walk through the setup and deployment of containerized LLMs on GPU node pools and see how KAITO can help reduce operational burden of provisioning GPU nodes and tuning model deployment parameters to fit GPU profiles.

5:00 PM - 6:30 PM (UTC)

In this session, we will walk through improving our understanding of how our intelligent apps are performing using the Azure Monitor Managed Service for Prometheus and Azure Managed Grafana. Additionally, we will improve our understanding of what our services cost, allowing us to find optimization opportunities, improve anomaly detection, and generally build a better understanding of how our services work together, all using the AKS Cost Analysis add-on.

4:00 PM - 5:30 PM (UTC)

Once we have our intelligent application running, it’s time to make it scale. Using globally-scalable services like Azure CosmosDB and Azure Front Door, along with AKS Fleet Manager, we can take bring our intelligent apps closer to their consumers, while providing a consistent operating experience for the developers and operations personnel responsible for the service.