I had the pleasure of being one of the four panelists at the recent Boston Azure AI Meetup (Virtual), where we discussed key takeaways from Microsoft Build 2025. My focus areas were NLWeb and the latest innovations in Azure AI Search.
Introducing NLWeb
NLWeb is a powerful open-source project by Microsoft that brings the capabilities of Large Language Models (LLMs) directly into your website. It enables intelligent, agent-compatible experiences over your structured item catalog with features like:
- Model Agnosticism – Easily switch out models, vector databases, and embedding strategies.
- Structured Data Support – Tools for importing your catalog and defining site scopes.
- Integration with Microsoft Copilot Platform (MCP) – Making your site ready for agentic interactions.
Get started with the GitHub repo and explore hands-on tutorials for setup, local deployment, and Azure hosting.
What’s New in Azure AI Search?
Azure AI Search is evolving to support richer, more intelligent search experiences:
- Agentic Retrieval Pipeline – Supports ad-hoc chunking, multimodal data, and RAG-ready ingestion workflows.
- Enterprise-Grade Security – Includes Entra-based access control and private preview of sensitivity labels via Microsoft Purview.
- Multimodal Demos – Now available via open-source GitHub repositories, showcasing the integration of text, image, and structured data.
🔗 Explore the updates:
Final Thoughts
Whether you’re building a smarter website or integrating AI-powered search into your enterprise app, both NLWeb and Azure AI Search provide developers with tangible tools and best-in-class infrastructure. You can now accelerate the building of intelligent, secure, and agile digital experiences.
Feel free to check out my LinkedIn or personal site for more updates, and join our Nashua Cloud .NET User Group for future meetups and learning sessions!