Vector databases like Qdrant and Milvus are specifically designed to efficiently store, manage, and retrieve embeddings. However, many applications already use relational databases like SQL Server or SQL Azure. In such cases, installing and managing another database can be challenging, especially since these vector databases may not offer all t...

I'm happy to share a new GitHub repository with developers: azure-sql-db-aspire. This collection of eight hands-on examples demonstrates how to integrate SQL Server and Azure SQL with .NET Aspire, making it easier to build modern, cloud-native applications. Why .NET Aspire for SQL Server? .NET Aspire is a new opinionated, cloud-ready stack for .N...

Security is a significant topic today, and the ability to access a service requiring authentication without using an API key, password, or secret is a common request from those concerned about the security of a solution, which includes all of us. In today's digital landscape, cybersecurity threats are increasingly sophisticated and frequent, mak...

In the previous article of this series, it was discussed how embeddings can be quickly created from data already in Azure SQL. This is a useful starting point, but since data in a database changes frequently, a common question arises: “How can the vectors be kept updated whenever there is a change to the content from which they have been generated?...

Embeddings and vectors are becoming common terms not only for engineers involved in AI-related activities but also for those using databases. Some common points of discussion that frequently arise among users familiar with vectors and embeddings include: Let’s tackle each one of these questions one by one starting from the very...

The RAG (Retrieval Augmented Generation) pattern, which is commonly discussed today, is based on the foundational idea that the retrieval part is done using vector search. This ensures that all the most relevant information available to answer the given question is returned and then fed to an LLM to generate the final answer. While vector search...

Project Mission The way people work and manage information is changing rapidly in our digital age. More and more people are struggling to keep track of all the online resources they use daily. They need a better way to save, organize, and retrieve important information from websites, articles, and other online sources. This is especially true fo...

Since the release of vector preview, we've been working with many customers that are building AI solution on Azure SQL and SQL Server and one of the most common questions is how to support high-dimensional data, for example more than 2000 dimensions per vector. In fact, at the moment, the vector type supports "only" up to 1998 dimensions for an emb...

I recently presented at Python Day 2024 on Langchain integration. I created a slide deck that I believe can be useful beyond just the session, so I wanted to share it here for everyone's benefit. The deck covers the most common topics for a Python developer: The slide deck, which I created using the Sli.dev SDK, so t...

Data is at the heart of every AI application, and efficient data ingestion is critical for success. With over 1,400 enterprise connectors, Logic Apps offers unmatched access to a diverse range of systems, applications, and databases, whether hosted in the cloud or on-premises. These connectors give businesses the flexibility to keep their data wher...