Konstantinos is a passionate IT professional and Microsoft Certified Trainer with over two decades of experience in crafting and deploying comprehensive IT solutions. His expertise lies in Microsoft technologies, evident in his Microsoft Azure MVP status. As a Cloud Solutions Architect, he specializes in designing enterprise architectures for cloud environments, including Azure, M365, and hybrid/multi-cloud solutions. Konstantinos is a lifelong learner, continually expanding his knowledge and certifications across various vendors, primarily Microsoft.
He is a dedicated community leader and advocate, spearheading knowledge-sharing initiatives through his Microsoft Tech Group, live sessions, and blog posts.
Konstantinos is a passionate IT professional and Microsoft Certified Trainer with over two decades of experience in crafting and deploying comprehensive IT solutions. His expertise lies in Microsoft technologies, evident in his Microsoft Azure MVP status. As a Cloud Solutions Architect, he specializes in designing enterprise architectures for cloud environments, including Azure, M365, and hybrid/multi-cloud solutions. Konstantinos is a lifelong learner, continually expanding his knowledge and certifications across various vendors, primarily Microsoft.
He is a dedicated community leader and advocate, spearheading knowledge-sharing initiatives through his Microsoft Tech Group, live sessions, and blog posts.
In this demonstration, we delve into building an Azure Container Apps stack. This innovative approach allows us to deploy a Web App that facilitates interaction with three powerful models: GPT-4, Deepseek, and PHI-3. Users can select from these models for Chat Completions, gaining invaluable insights into their actual performance, token consumption, and overall efficiency through real-time metrics.
This deployment not only showcases the versatility and robustness of Azure AI Foundry but also provides a practical framework for businesses to observe and measure AI effectiveness, paving the way for data-driven decision-making and optimized AI solutions.
The proposed workshop is a practical, hands-on demonstration focused on deploying a secure, scalable, and cost-effective multi-model generative AI solution using Azure AI Foundry Serverless Models. The core of the solution is a multi-model web application—the Azure AI Foundry Multi-Model Tool—that allows users to interact with multiple models like GPT-4, DeepSeek, and Phi-3 via serverless API deployments. Participants will gain practical insights into modern serverless AI architecture, including how the frontend (Vite + React) and backend (Python + FastAPI) are securely integrated using Azure Container Apps, Dapr for service-to-service communication, and Azure Key Vault for secret management. The content and exercises will emphasize comparing models by performance and efficiency (e.g., tokens/sec, total time) and ensuring a highly available, secure, and cost-effective deployment.
Understand the architecture and benefits of serverless AI deployment on Azure AI Foundry.
Deploy a multi-model solution using Azure Container Apps, Dapr, and Azure Key Vault.
Build the backend with Python + FastAPI to manage model selection and API integration with Azure AI Foundry.
Evaluate and compare the performance and efficiency of different large language models (e.g., GPT-4, DeepSeek, Phi-3) in a real-time web application.
Implement a secure secret management strategy using Azure Key Vault and Managed Identity
AI Software Developers, Cloud Architects and Technical Decision-Makers, Data Scientists and AI Engineers, Developers looking to build scalable and secure generative AI applications, Professionals interested in comparing LLM performance and optimizing deployment costs
Own laptop. An active Azure subscription. Azure AI Foundry Hub with an active project (e.g., in East US region). Basic knowledge of Python and web development concepts (e.g., REST APIs, frontend/backend basics). VSCode with the Azure Resources extension installed. Familiarity with foundational AI concepts and Azure services is recommended.
The complete multi-model web application code (Frontend: Vite + React; Backend: Python + FastAPI). Instructions and files (e.g., Dockerfile, requirements.txt) to deploy the solution on Azure Container Apps. The presentation/blog post presented during the workshop.
A discussion of building global success and how artificial intelligence is transforming the world.
