.NET Rocks!

How will coding agents change your code? While at Build, Carl and Richard chatted with Scott Hunter about the announcements around coding agents at the keynote. Scott talks about the agent mode available in Visual Studio Code - and now in Visual Studio! Agent mode allows the LLM to evaluate the code across an entire solution, not just the file you're currently looking at. You can create a workflow where GitHub issues are assigned to the agent, which then generates code and provides a pull request for evaluation. The agents are here and helping us do more!

Ken Exner, the Chief Product Officer at Elastic, shares insights on groundbreaking serverless storage and search capabilities. He highlights the financial benefits, emphasizing the pay-as-you-go model for data storage and movement. The conversation dives into Elastic's integration with Azure AI Foundry, showcasing its potential for providing rapid access to data for large language model implementations. Ken also discusses the importance of vector databases, hinting at their vital role in advanced hybrid search techniques and the future of AI applications.

Chris Woodruff, a seasoned software developer and Microsoft MVP, shares his expertise on C# networking. He dives into the fundamentals of networking, exploring options like web sockets, gRPC, and SignalR. The conversation also touches on the upcoming QUIC standards in HTTP/3, promising to simplify networking. Listeners can learn about the evolution of technology through personal anecdotes and challenges faced in the early days of internet connectivity, making the topic both informative and relatable.

Erwin van der Valk, a Principal Engineer at Duende Software, dives into the intricacies of securing browser-based frontends with ASP.NET Core backends. He shines a light on the Backend for Frontend (BFF) Security Framework, discussing Sam Newman's BFF Pattern to manage diverse clients. The conversation highlights the complexities of OAuth 2.0 in tackling backend security, and Erwin emphasizes the necessity of layered security measures, cookie management, and efficient workflows to protect user identity and enhance application safety.

Steve Smith, an entrepreneur and software developer dedicated to building quality software, shares valuable insights on balancing architecture and code. He discusses how architectural strategies evolve with team size and the impact of modern tools on decision-making. The conversation also delves into the challenges of refactoring, the importance of simplicity in coding practices, and how to maintain project direction in software development. A compelling dialogue that highlights the delicate dance between effective architecture and efficient coding!

Rob Mensching, CEO and co-founder of Fire Giant and creator of WiX, shares insights on the burnout crisis facing open source maintainers. He proposes an innovative maintenance fee model through GitHub to support these developers sustainably. The conversation highlights the challenges of single maintainers overwhelmed by user demands and emphasizes the importance of fair compensation. Rob discusses how this model can help balance the community’s needs with the realities of maintaining critical software projects.

Vishwas Lele, Co-founder and CEO of pwin.ai and Microsoft Regional Director, discusses how AI can revolutionize responses to government RFPs. He shares insights from his startup journey, emphasizing the importance of tuning large language models to maximize their effectiveness. The conversation covers the need for domain-specific solutions and highlights the balance between speed and quality in proposal writing. Vishwas also addresses the challenges of using AI in document processing, stressing transparency and compliance in RFP responses.

Elias Puurunen, an innovator in digital event experiences and founder of Tractus Events, dives into integrating audio-video in .NET applications. He explains the benefits of using NDI protocols for real-time streaming and the power of P/Invoke for accessing libraries. The conversation highlights the use of NVIDIA GPUs for swift encoding and decoding, showcasing advancements in video production tools. Elias also emphasizes the evolution of video compression techniques and how modern frameworks enhance performance for live streaming.

In this discussion, Spencer Schneidenbach, a Microsoft MVP and CTO of Aviron Software, shares his expertise on integrating large language models in healthcare applications. He elaborates on using AI to improve customer interactions by summarizing conversations and analyzing sentiments. The chat touches on Microsoft's Semantic Kernel, which simplifies connecting AI to APIs. Additionally, Spencer highlights the balance between creativity and consistency in AI responses, emphasizing the necessity of effective prompting and rigorous testing for successful AI implementations.

Dr. Jodie Burchell, a developer advocate in data science at JetBrains and former lead data scientist at Verve Group Europe, discusses measuring large language models (LLMs). She dives into various benchmarks and the importance of accuracy, reliability, and customization for specific topics. The conversation highlights the challenges in building effective test suites and emphasizes that smaller, targeted models can often outperform larger counterparts. Jodie also explores the complexities of evaluating AI performance with humor and insight.