Browse Machine Learning Videos (48)

Paula Santamaría and Julia Schröder Langhaeuser present a production Retrieval-Augmented Generation (RAG) architecture built on Azure Database for PostgreSQL, explaining why Postgres can be a solid foundation for RAG at scale and what it takes to move from prototype to production with performance tuning and monitoring.

Microsoft Developer previews Livestream 3 of POSETTE 2026 with short introductions to 11 PostgreSQL talks, including WAL, logical replication, testing and coverage, consistency in clusters, performance tuning, and a session on production RAG at scale using Azure Database for PostgreSQL.

Mohsin Ejaz explains how to build safety tooling and guardrails for automated, AI-driven PostgreSQL tuning, focusing on monitoring, validation, and risk controls so performance improvements don’t come at the cost of outages or regressions.

dotnet explains patterns for modernizing data and migrating line-of-business applications incrementally, focusing on moving the database first and evolving the app through stable API layers and modern data access approaches.

John Savill rounds up a week of Azure platform changes and retirements, spanning compute/storage updates, database and identity improvements, monitoring changes, and several developer-facing AI items including GitHub Copilot Agent Mode in SSMS and Azure AI Foundry agent licensing and model availability.

Laurent Bugnion explains why Microsoft Build 2026 labs are a popular way for developers to learn through hands-on sessions, including tracks that cover AI, Copilot, and Microsoft Fabric. He also shares where to find the on-demand “Digital lab” sessions and how long they remain available after the event.

Dan Wahlin demonstrates an “agentic journey” workflow that takes an app idea through planning, coding, infrastructure creation, and deployment to Azure, using GitHub Copilot CLI and Azure skills to handle tasks like Bicep templates, health probes, and database wiring for an app backed by Azure SQL and Microsoft Foundry.

John Savill runs through a Build-special weekly Azure update, covering a wide set of platform announcements across compute, containers, integration, monitoring, databases, Fabric/Databricks, and Azure AI Foundry—plus security-focused items like confidential computing and Purview agent integrations.

Charles Feddersen and Abe Omorogbe explain how AI apps and agents change database design, focusing on reasoning over operational data instead of only transactions. They demo new capabilities across Azure SQL Database, Azure Cosmos DB, and Azure HorizonDB (cloud-native PostgreSQL) to simplify architectures and reduce latency.

Milos Colic shares how Xoople scaled Python-based AI workloads on Azure using Ray via Anyscale, covering the distributed-systems challenges behind data ingestion, training, and inference, and why the team prioritized delivering outcomes over operating clusters.

Nikisha Reyes-Grange introduces Azure HorizonDB and Rayfin, focusing on how these Azure Data and Microsoft Fabric innovations aim to modernize PostgreSQL operations and simplify building and running data applications, including SQL-level AI functions and hybrid search concepts.

Marco Casalaina and Ayca Bas explain how to build grounded, enterprise-ready agents by delegating context to three “IQ” systems: Foundry IQ for knowledge, Fabric IQ for data, and Work IQ for human/work context, with a live demo showing how this approach avoids brittle, hand-wired pipelines.

Microsoft Developer’s Data Exposed episode shows how to build a data-powered application using Rayfin with a Microsoft Fabric SQL Database backend, including Fabric SSO authentication. It also covers iterating on the app with GitHub Copilot and how Rayfin’s code-first SDK reduces the amount of infrastructure wiring you need to do.

Amanda Silver and Marco Casalaina explain how to build context-aware AI agents by connecting them to enterprise knowledge, business data, and work signals using Foundry IQ, Fabric IQ, and Work IQ, with an emphasis on orchestration, governance, and operating within trusted boundaries.

Mark Russinovich and Ion Stoica discuss how AI platforms need to evolve for agentic, multimodal, globally distributed workloads, covering infrastructure fundamentals, training and real-time serving architectures, and why open source, security, and governance are becoming core platform requirements.

Sunitha Muthukrishna demonstrates how to use Rayfin with Microsoft Fabric to generate and deploy an agent-driven full-stack web app, including a managed database, authentication, and backend services, then connect it to an existing Fabric data estate to add analytics, BI, and AI-powered experiences.

Alexander Wachtel shows how Microsoft Fabric can be used to build dynamic, context-aware multi-agent workflows, using Fabric foundations like semantic models, lakehouses, and pipelines to support planning, retrieval, summarization, and execution across specialized agents.

Rishabh Saha shares how Microsoft and PepsiCo engineers modernized PepsiCo’s data foundation for agentic applications, using Azure SQL, Cosmos DB, PostgreSQL, and Azure Databricks. The session outlines a practical build path for agentic RAG, including Azure SQL vector indexing and semantic search to speed up repeatable app patterns.

Rob Ferguson leads a Microsoft Build 2026 panel on shipping custom AI models at scale, covering practical trade-offs in fine-tuning and serving, plus what teams are doing to control inference cost and latency in production.

Ben Zulauf and Sachin Patney present a Microsoft Build 2026 session on Rayfin’s approach to building backends as code, with agent-driven app generation and native integration with Microsoft Fabric data and analytics for deploying secure, scalable applications.

Parthasarathy Srinivasan and Rajya Laxmi Yellajosyula demonstrate a multi-cloud, enterprise AI workflow that combines Oracle Database@Azure with Microsoft Fabric, MCP, and GitHub Copilot, covering provisioning, synthetic data creation, ETL from bronze to gold, and an end-to-end fraud detection demo driven by natural-language orchestration.

Pablo Castro presents a Microsoft Build 2026 deep dive into Foundry IQ, Microsoft’s context engineering platform for building agents that can retrieve enterprise knowledge using agentic RAG. The session covers Foundry IQ’s architecture, connecting new knowledge sources, ingestion pipeline customization, retrieval APIs, and performance/evaluation improvements.

Vaidyaraman Sambasivam, Osi Otugo, and Jean Boudier demonstrate an end-to-end flow for taking Hugging Face open-source models from discovery to production inference using Foundry Managed Compute in Azure AI Foundry, focusing on scaling, governance, and avoiding direct GPU management.

William Liang demonstrates how teams use Azure AI Foundry to distill large models into smaller, task-focused language models using supervised fine-tuning, with an emphasis on reducing production latency and cost while maintaining accuracy through structured evaluation.

Katarina Stanley and Daniel Arrizza explain how Anyscale on Azure uses Ray on Azure Kubernetes Service (AKS) to run distributed AI workloads, from multimodal data pipelines and training/fine-tuning through to deploying models as inference services inside an Azure subscription.

Dave Citron (CVP, Microsoft AI) walks through what goes into training Microsoft’s latest MAI model family—covering new thinking, coding, voice, transcription, and image models, plus the architectural and evaluation choices behind their capabilities and performance.

Tessa Kloster, Arindam Chatterjee, and Anshul Sharma present a Microsoft Build 2026 session on using Microsoft Fabric Real-Time Intelligence to build event-driven AI applications and autonomous agents that react to live data, combining streaming ingestion, real-time analytics, and actioning in a governed workflow.

Kim Manis explains how Microsoft Fabric supports a secure and scalable data estate, covering governance with OneLake Catalog, compliance integration with Microsoft Purview, capacity controls, and developer workflows like Terraform and the Fabric CLI, plus how these foundations enable grounded AI agents with Foundry and OneLake.

Yohan Lasorsa demonstrates how Rayfin on Microsoft Fabric provides a code-first backend with type-safe schemas, APIs, functions, storage, and hosting, taking an app from idea to deployment while keeping Fabric-native data ready for governance, analytics, and AI use from day one.

Jeff Smith and Ram Kakani show how Oracle managed MCP Servers can connect Oracle Database@Azure to Microsoft IQ (Work IQ, Fabric IQ, and Foundry IQ) so teams can build agentic, AI-driven workflows with more context, reasoning, orchestration, and governance over enterprise data.

Andrew Leader and Maha Bayana explain how Windows ML enables local AI apps on Windows using custom or open-source ONNX models, with a focus on running inference efficiently across CPU, GPU, and NPU. They also cover what’s new, including WebNN support for web scenarios and improved tooling via AI Toolkit for VS Code.

Thomas Maurer talks with Kyle Ikeda about the Microsoft Agent Pre-Purchase Plan and how Agent Commit Units (ACUs) can be used across Azure AI Foundry, Copilot Studio, Fabric, and GitHub to make AI agent spend more predictable, with guidance on purchasing, tracking consumption, and budgeting responsibly.

Microsoft Developer explains how to add persistent memory to an AI agent using Microsoft Agent Framework, storing conversation history in SQL Server and showing how the same approach can be future-proofed with Azure SQL Database.

Elvis Kahoro joins GitHub’s Open Source Friday to explain dlt (from dltHub), an open source Python SDK for building production-grade data pipelines, including what it is, who it’s for, and how it helps developers move data without getting buried in pipeline complexity.

Microsoft Developer introduces STATE-Bench, an open-source benchmark for evaluating whether “memory” actually improves AI agent performance on realistic, stateful enterprise tasks, focusing on workflow execution quality, repeatability, efficiency, and user experience rather than simple recall tests.

Authorised Territory demonstrates a .NET data ingestion pipeline that converts a PDF to Markdown via the MarkItDown MCP server, generates embeddings with a local Ollama model, and stores those embeddings in SQL Server 2025 running in Docker Desktop.

John Savill explains why enterprises need a data virtualization layer and how to build one using Microsoft Fabric OneLake, including a single namespace approach, shortcuts, mirroring, governance, and semantic models to make data easier to use for analytics and AI.

John Savill's Technical Training gives a quick overview of why enterprises often need a data virtualization layer, and how it helps provide a unified way to access data across different systems.

Microsoft Developer recaps key themes from Azure Cosmos DB Conf 2026, focusing on what engineers are doing in production to scale reliably, keep costs under control, and support AI-driven workloads—especially vector search and modern search patterns built into Cosmos DB.

John Savill breaks down practical ways to change an AI model’s behavior, from prompt and context techniques through to retrieval-augmented generation (RAG) and fine-tuning approaches like LoRA.