The Role of LLMs in Visual Studio Productivity with Leslie Richardson - .NET Rocks! Recap

Podcast: .NET Rocks!

Published: 2025-11-20

Duration: 1 hr 1 min

Guests: Leslie Richardson

Summary

Leslie Richardson discusses how large language models (LLMs) are enhancing productivity in Visual Studio by integrating into debugging tools and simplifying code generation, stressing the importance of discoverability and usability for developers.

What Happened

Leslie Richardson, a product manager on the .NET and C# developer experience team, delves into the integration of large language models (LLMs) within Visual Studio to enhance debugging and productivity. She highlights how these models can assist with generating link queries and analyzing code, making the debugging process more intuitive and less reliant on manual intervention. Leslie discusses her experience with debugging tools when she joined Microsoft and her initial apprehension about working on the debugger due to her limited experience with it at the time.

The conversation covers the evolution of Visual Studio and Visual Studio Code, with Leslie noting the balance between offering powerful features and maintaining a user-friendly interface. She reflects on the challenges of making lesser-known features more discoverable, such as pinnable properties, which aim to make debugging more accessible.

Leslie emphasizes the role of AI in suggesting and aiding in code improvements, likening it to an assistant that can reveal the depth of available debugging tools that many developers might not be aware of. The introduction of Copilot in Visual Studio 2026 is discussed, which aims to provide contextual help and automate repetitive tasks, thus enhancing developer productivity.

The episode also touches on the potential of AI to transform coding practices, with tools like Copilot enabling developers to write tests more efficiently and explore features they might not have previously utilized. Leslie shares insights into the development of new features for Visual Studio, such as the enumerable visualizer, which leverages Copilot to help developers better understand and manipulate data structures.

A significant portion of the discussion revolves around the use of Microsoft Command Palette (MCP) servers and how they can make complex tasks, like writing Kusto queries, more manageable through natural language processing. Leslie explains how she uses MCP servers for simplifying Git operations and enhancing her workflow.

The episode concludes with thoughts on how Visual Studio and Visual Studio Code cater to different developer needs, balancing between offering comprehensive toolsets and maintaining simplicity for new developers. Leslie's insights underscore the ongoing evolution of development environments, driven by advancements in AI and user feedback.

Key Insights