--- title: "Create multimodal AI agents with persistent memory | DEMSP390" url: https://daily.dev/posts/create-multimodal-ai-agents-with-persistent-memory-demsp390-q0qoa7okh source_url: https://www.youtube.com/watch?v=o_rcIQUU-k4 type: video:youtube source: "Microsoft Developer" published: 2026-06-04T14:08:43.072Z updated: 2026-06-04T14:09:17.776Z tags: ["ai-agents", "mcp", "multimodal"] reading_time: 19 upvotes: 1 comments: 0 language: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Create multimodal AI agents with persistent memory | DEMSP390 **[Microsoft Developer](https://daily.dev/sources/microsoftdeveloper)** · 19 min read · 1 upvotes · 0 comments ## Summary A conference talk from Microsoft Build showcasing Napster's Omni Agent API, a platform for creating multimodal AI agents with persistent memory. The system integrates with Azure AI Foundry and uses an 'edge MCP' server embedded in website JavaScript to let agents control web apps without VLM screenshot analysis. Agents support video, audio, text, WhatsApp, and phone channels. A key technical innovation is auto-generating an MCP server from existing source code at authoring time, enabling a single developer to deploy a fully functional multimodal agent in one day. The platform is available as an Azure native offering with unified billing, and video avatar costs have been reduced to 1 cent per minute. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#mcp](https://daily.dev/tags/mcp), [#multimodal](https://daily.dev/tags/multimodal) [View this post on daily.dev](https://daily.dev/posts/create-multimodal-ai-agents-with-persistent-memory-demsp390-q0qoa7okh)