Microsoft Events hosts an advanced Ignite session featuring Mark Brown and Jai Maldonado, demonstrating how to use Cosmos DB within Microsoft Fabric to build AI-enabled, real-time applications at scale.

Real-Time Analytics and AI Applications with Cosmos DB in Microsoft Fabric

Speakers: Mark Brown, Jai Maldonado
Session: BRK228 — Microsoft Ignite 2025

Overview

Discover how Azure Cosmos DB and Microsoft Fabric simplify the creation of AI-enabled, real-time applications. This session provides a deep dive into:

• Scaling out architectures for low-latency performance
• Integrating across the Microsoft ecosystem, including OpenAI-powered scenarios
• Processing customer data for real-time intelligence
• Leveraging Apache Spark for machine learning and recommendation engines
• Using vector search in Cosmos DB for high-performance personalization
• Employing collaborative filtering algorithms to enhance recommendations
• Managing blue-green deployments for safe application updates
• Streamlining data access without additional middleware
• Building and shipping real-time solutions using Notebooks and built-in user functions

Key Topics Covered

1. Scale-Out Architecture and Performance

• How Cosmos DB supports low-latency, scale-out workloads
• Architecture design to handle large-scale, real-time data

2. Integration and Ecosystem

• Real-world integration scenarios with Microsoft Fabric
• Case study: Incorporating OpenAI models for enhanced analytics

3. Real-Time Data Processing

• Using Real-Time Intelligence (RTI) to process and analyze customer data as it arrives
• Demonstration: Real-time analytics without manual ETL using Fabric Zero ETL features

4. Advanced Analytics and Machine Learning

• Implementing collaborative filtering for personalized recommendations
• Training ML-based recommendation engines with Apache Spark
• Seamless workflow using Fabric Notebooks for experimentation and testing

5. Vector Search and Personalization

• Combining vector search capabilities in Cosmos DB to deliver high-performance, personalized user experiences

6. Deployment and Data Functions

• Blue-green deployment strategies for robust AI apps
• Utilizing user-defined functions to customize data transformations

7. Interactive Reporting and Pricing Analysis

• Creating interactive reports to visualize correlations between pricing and customer reviews
• Strategies for dynamic price analysis and customer engagement

Resources

• Microsoft Ignite 2025 Sessions
• Cosmos DB in Fabric: Real-Time Intelligence Plan

Summary

This session is ideal for data engineers, architects, and developers seeking to harness Microsoft Fabric and Cosmos DB for the next generation of AI-driven, real-time applications, enabling scalable analytics, personalization, and robust deployment architectures.