At the Microsoft Build conference, Fabric, PostgreSQL and Cosmos DB received AI enhancements

Microsoft recently made about 60 announcements at the Build conference, including new artificial intelligence capabilities for its cloud database management products.

Fabric, the company’s unified data platform launched last year, is a major beneficiary. A Workload Development Kit, currently in preview, can be used to extend applications in Fabric. Fabric Data Sharing is a new feature that handles real-time data across users and applications. It includes an application programming interface for accessing data stored in external sources. New automation capabilities (Automation) simplify repetitive tasks.

Fusion is a new RESTful GraphQL API that allows Fabric developers to access multiple data from different sources through a single query. Expanded user data capabilities enable building data-centric applications in Fabric lake libraries, data warehouses, and mirrored databases with simple integration using native code capabilities and custom logic.

Microsoft said that Fabric has added generative artificial intelligence capabilities to artificial intelligence, allowing non-technical users to build applications that can answer questions in natural language. The company also added Azure OpenAI Services at each layer to create data flows and pipelines, generate code, and build machine learning models.

A new real-time intelligence feature in Fabric can be a software-as-a-service application that creates a single place to ingest, process and route events from disparate sources. Event streams can be processed using preconfigured streaming connectors and connected to cloud sources via content-based routing. This real-time intelligence feature can be flexibly configured to suit different business needs. It can connect to cloud sources based on internal routing, and can also handle pre-configured stream connectors for event processing. This flexibility allows it to receive, process and route events from different sources in real time. At the same time, it can connect to cloud sources through content-based routing to ensure that events are processed correctly according to business needs. With this new real-time intelligence capability, Fabric can provide

Fabric Copilot in private preview can be used to generate queries to detect unknown conditions in large-scale data that are beyond the scope of human analysis.

Artificial Intelligence for PostgreSQL, CosmosDB developers

Fine-tuning for PostgreSQL enables access to artificial intelligence capabilities through Azure OpenAI services or in-database models. This provides an option for those who wish to keep their data within the database instance.

The Azure+ AI extension lets developers leverage Azure AI’s large language models in their PostgreSQL applications. They can call the Azure OpenAI service to generate LLM-based vector embeddings to achieve efficient similarity search, and they can also call Azure AI Language to handle scenarios such as sentiment analysis, language detection, and entity recognition.

Developers can also invoke pre-trained machine learning models for scenarios such as fraud review, detection and product recommendations. Real-time text translation can be done using Azure AI Translator.

The in-database embedding generation function supports the text embedding model in Azure Database for PostgreSQL, and the embedding can be generated within the database without calling Azure OpenAI Service. Microsoft says this reduces embed creation time to a few milliseconds of latency and makes the cost more predictable.

CosmosDB, the globally distributed multi-model database service for building large-scale applications, is getting several artificial intelligence-related updates. Cosmos DB for NoSQL now has built-in vector indexing and vector similarity search, keeping data and vectors in sync without the need for a separate database. This feature is provided by DiskANN, an open source software suite of approximate nearest neighbor search algorithms, and is currently in preview.

A new feature now in preview allows users to transition their serverless Azure Cosmos DB accounts to provisioned capacity mode through the Azure portal or command line interface while retaining full access to data operations.

A new option in preview lets Cosmos DB for MongoDB users create a continuously updated copy of the cluster in another region for failover. The new Go software development kit can operate databases, containers and projects across multiple regions to achieve high-availability applications.