Scale enterprise .NET apps with AI‑assisted cross‑platform workflows | DEMSP394
Sam Basu and Colin Whitlatch demonstrate an end-to-end, AI-assisted workflow for scaling enterprise .NET apps across web, desktop, and mobile using Uno Platform Studio, focusing on contextual AI that can inspect live app state and help iterate on UI and app behavior.
Overview
This Microsoft Build 2026 session focuses on enterprise .NET development scenarios (layered systems, cross-platform targets, mixed-skill teams, and production constraints) and shows how an AI-assisted workflow can support real-world app iteration beyond greenfield code generation.
Session flow (from the published chapters)
Overview of the software development spectrum and AI’s role
- Positions AI tooling in the broader development workflow, with emphasis on enterprise constraints and the need for context-aware assistance.
Uno Platform Studio launch and productivity features
- Introduces Uno Platform Studio and highlights productivity-oriented capabilities aimed at cross-platform .NET app development.
Creating a sample weather app using an AI prompt
- Demonstrates generating or scaffolding a sample app via prompting as part of the workflow.
Design mode and live visual tweaking in the browser
- Shows UI iteration using a design mode experience with live visual adjustments.
Desktop app run: hot reload and live design editing
- Demonstrates running a desktop app and iterating quickly using hot reload and live design editing.
Terminal mode for headless, fast operations
- Switches to a terminal-oriented workflow for faster, headless operations.
Launching a desktop app using AI for non-deterministic project setup
- Demonstrates using AI assistance during project setup steps that can be non-deterministic or context-dependent.
Reviewing the completed weather app built fully in the browser
- Reviews the resulting app and the browser-based workflow end state.
Preview and validation tools for UI components and snippets
- Shows tooling for previewing and validating UI components/snippets as part of the iteration loop.