ISE Developer Blog
Solving global tech challenges, sharing insights, and empowering developers
Latest posts
Leveraging Fabric Eventhouse to Store Great Expectations Validation Results
A step-by-step guide to storing data validation results into the Eventhouse & visualize data using queries and dashboard.
Taming Complexity: Intuitive Evaluation Framework for Agentic Chatbots in Business-Critical Environments
This blog post introduces a comprehensive evaluation framework for enterprise chatbots powered by large language models (LLMs), specifically addressing the challenges of assessing Line of Business (LOB) agents in business-critical environments. The authors tackle the fundamental problem that traditional chatbot evaluation metrics fail to capture the nuanced, non-deterministic performance of modern LLM-based systems, proposing a solution that combines realistic chat simulation using an LLM-powered User Agent, automated ground truth generation at scale, and comprehensive metrics including function call precision, r...
Unlocking Vector Search with OneLake Indexer and OpenAI Integration in Microsoft Fabric
Exploring how Microsoft Fabric OneLake indexer integrates with OpenAI
Ground Truth Curation Process for AI Systems
Steps to Produce High Quality Ground Truth Pairs for AI Systems
Customizable and Extensible Evaluation of a GenAI Application
Suggestions on how to evaluate a generative AI application in an adaptable and extensible manner.
Learnings from External Data Handling
This blog post discusses the challenges and solutions encountered by the ISE team at Microsoft while making a distributed system production-ready. It focuses on issues including slow processing speeds and out-of-memory exceptions, and provides insights into the methods used to address these problems.
Enhancing Analytical Queries with Semantic Kernel and In-Memory SQL Processing
Integrating Semantic Kernel with an ephemeral in-memory SQL plugin, enabling secure and efficient analytical queries over structured data.
AI Model Promotion with dstoolkit-mlops-v2
Evaluates various repository structures and designs for maximizing the efficiency of Data Scientists and Software Engineers developing, promoting and deploying AI models on the same project.
Use Agent to Update Dataverse Table Content
Create an agent using Copilot Studio to update the content of a Dataverse table.