Microsoft Build 2024: New AI & developer tools

June 13, 2024 Scott Hermes

Blog

Event highlights

Microsoft Build 2024 has concluded, leaving some 200,000 developers and tech enthusiasts buzzing with excitement about groundbreaking innovations, particularly in AI and developer tools, after three days of immersive learning, skill-building, and community connection.

This annual flagship event included 300+ sessions, demos, and expert-led labs available on-demand. Here are the highlights!

Featuring Copilot+ PCs, Copilot Stack, Microsoft Copilot

Copilot+ PCs

Microsoft unveiled Copilot+ PCs, equipped with a brand-new AI-chip (NPU). These are touted as the fastest, most intelligent PCs ever released by the company, promising enhanced performance and smarter functionalities.

The Copilot Stack

This advanced toolset empowers developers to build their own copilots. The Copilot Stack incorporates Microsoft's extensive learnings in AI to provide a robust foundation for creating intelligent assistants.

Microsoft Copilot

Positioned as your everyday AI companion, Microsoft Copilot integrates seamlessly into various applications, enhancing productivity and user experience with intelligent assistance.

Day 1 Highlights

AI security session with Mark Russinovich

Azure CTO Mark Russinovich delivered an insightful session on AI security. He emphasized the potential threats posed by large language models (LLMs) and underscored the importance of securing these technologies against malicious exploitation.

GPT-4 in Azure OpenAI service

Although previously announced by OpenAI, the availability of GPT-4 on Azure was a major highlight. This latest model boasts impressive capabilities, including voice recognition, speech functionalities, and the ability to handle multi-turn conversations with ease.

Real-time intelligence in Fabric

Microsoft announced the release of EventHouse, a feature within Fabric that incorporates real-time data and analysis. This enhancement ensures Fabric remains a comprehensive, user-friendly data platform, eliminating the complexities of resource management.

Day 2 highlights

AKS Automatic simplifies setup process

Microsoft introduced AKS Automatic, an enhanced version of Azure Kubernetes Service. This new deployment option simplifies the setup process by integrating best practices for security, networking, and maintenance, allowing developers to deploy production-level workloads swiftly.

GitHub Copilot for Azure eases development

A significant update to GitHub Copilot now includes integration with Azure. By typing @azure in the Copilot chat, developers can gain insights into their environment and generate Infrastructure as Code for immediate deployment, streamlining the development process.

Copilot Studio gets host of new capabilities

Microsoft’s low-code/no-code toolkit, Copilot Studio, received a substantial upgrade. Developers can now create virtual assistants with agents and triggers, enabling sophisticated workflows. This tool promises a 175% growth quarter over quarter, showcasing its increasing adoption and utility.

Live demo— building an AI assistant

Scott Hanselman and Mark Russinovich entertained attendees with a live demonstration of building an AI assistant. The session highlighted the nuances of working with LLMs, the use of Semantic Kernel, and provided a humorous yet educational look at AI development.

Day 3 highlights

Building a Copilot

Jeff Omhover and Sindhura Bandhakavi from the Azure AI studio team led an in-depth session on creating copilots. They discussed various aspects such as data selection, transitioning from proof-of-concept to production, and utilizing features like promptflow and AI studio for monitoring and evaluation.

End-to-end AI development with Prompty

Seth Juarez introduced the new prompt file standard, Prompty, for defining LLM interactions through markdown. Prompty offers a lightweight, open-source solution for orchestrating LLM calls, presenting a simpler alternative to more complex frameworks.

PII extract feature in Azure AI Studio

A noteworthy feature of Azure AI Studio is its ability to use LLMs to identify and redact Personally Identifiable Information (PII) in log messages. This not only enhances privacy but also adds a layer of amusement by making the logs appear humorously censored.

Build AI with Kin + Carta

Microsoft Build 2024 was a resounding success, packed with innovations and valuable insights. From AI advancements to enhanced developer tools, the event showcased Microsoft's commitment to driving the future of technology. Attendees left with a wealth of knowledge, equipped to implement these new tools and ideas in their own projects. As we look forward to next year’s event, the excitement and potential of what’s to come are palpable.

Have questions or want to learn more?

Contact our experts