Create multimodal AI agents with persistent memory | DEMSP390

From the creator

Your users reach you through your website, your app, your store, your support line. And whether they were in touch five minutes or five days ago, every channel starts from zero, every time. Multimodal AI agents with persistent memory close that gap, and you can have them up and running now. In this session, Napster CTPO Edo Segal builds a working video AI agent live using the Napster Omniagent API. You'll leave with the architecture, the code patterns, and a clear next step to build your own agent that lives across every surface your organization touches. 𝗦𝗽𝗲𝗮𝗸𝗲𝗿𝘀: * Edo Segal 𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻: This is one of many sessions from the Microsoft Build 2026 event. View even more sessions on-demand and learn about Microsoft Build at https://build.microsoft.com DEMSP390 | English (US) | Agents & apps Demo | (300) Advanced #MSBuild Chapters: 0:00 - Mention of book exploring AI’s impact on society and future generations 00:02:50 - Guide on how to start creating multimodal agents via API and vibe coding 00:05:14 - Walkthrough of provisioning Napster as an Azure Resource and using Omni Agent 00:05:48 - Partnership benefits enabling streamlined enterprise procurement 00:07:16 - Technical breakthrough: embedding MCP Server directly in JavaScript for local intelligence 00:09:27 - Paradigm shift enabled by frontier models 00:10:39 - Secure integration within Azure AI Foundry with minimal roles 00:15:18 - Real-world analogy: Replacing in-person assistants with digital Omni agent

Published 2026-06-04 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.