Back to Podcast Digest
dotnet46m

Taking your AI to the edge with .NET MAUI

TL;DR

  • Copilot is now embedded across the MAUI team's whole workflow: Gerald says the team uses AI not just for code generation but also for bug reproduction, test selection, review, documentation, and post-merge learning, with Copilot-authored merged PRs rising from 7% to 62% in under a year.

  • .NET MAUI 10 focused on stability, performance, and less XAML pain: The release adds global namespaces, implicit namespaces in preview, and XAML source generation that Gerald says can deliver up to 78,000% faster inflation and 99% less debug memory in some cases.

  • Safe area handling and Material 3 are practical upgrades developers will feel immediately: MAUI 10 introduces a unified safe-area API for iOS and Android layouts, plus opt-in Android Material 3 support rolling out through service releases and expected to be fuller by .NET 11.

  • .NET 11 will quietly make a big runtime shift: MAUI is moving from Mono to CoreCLR by default on iOS, Android, and Mac, which Gerald frames as a mostly smooth transition that brings MAUI closer to the rest of .NET tooling and diagnostics like dotnet-trace and dotnet-counters.

  • The new CLI and DevFlow are built for both developers and agents: MAUI is adding dotnet run, dotnet watch, and a DevFlow system that can inspect the visual tree, take screenshots, read sensor values, and let Copilot autonomously test app behavior in emulators and simulators.

  • Microsoft is pushing on-device AI as a first-class MAUI scenario: Through Microsoft.Maui.Essentials.AI, developers can call Apple Intelligence, Gemini Nano, Windows Copilot Runtime, ONNX, or cloud models through a unified API, with demos like David Ortinau's Barista Notes app using voice intent and image capture to count people in a room and recommend coffee shots.

The Breakdown

The .NET MAUI team says GitHub Copilot changed how they build the framework itself, pushing Copilot-authored merged PRs from 7% to 62% in under a year while they shipped major MAUI 10 improvements and previewed a very AI-heavy MAUI 11. Gerald's core pitch is simple: use .NET MAUI to ship native apps everywhere, then bring AI all the way to the edge with local models, agent-friendly tooling, and new app automation features like DevFlow.

Was This Useful?

Share