ISE Developer Blog
Solving global tech challenges, sharing insights, and empowering developers
Latest posts
Hierarchical Waterfall Evaluation of Query Classification, Retrieval & Generation in Multi-Agent LLM Systems
This post presents a hierarchical waterfall framework for evaluating query classification, retrieval, and generation in multi-agent LLM systems.
Efficient Ground Truth Generation for Search Evaluation
This article discusses the challenges and methodologies for creating a reliable groundtruth dataset for search evaluation. It highlights the use of TREC Pooling and GPT-4o assisted ranking to overcome the labor-intensive process of manual labeling, ensuring high-quality search evaluations even with limited data.
Durable Functions for Indexing in RAG: A Practical Python Approach
Azure Durable Functions streamline RAG indexing by combining push flexibility with pull reliability for scalable,
Cloud infrastructure provisioning: best practices for IaC
Infrastructure-as-Code (IaC) is a common DevOps practice that enterprises use to provision and deploy IT infrastructure. Enterprises applying IaC and continuous integration/continuous delivery (CI/CD) pipelines can maintain high availability and manage risk for their cloud environments at scale. However, multi-environment challenges, manual processes, and fragmented guidance still lead to configuration drift, errors, and inconsistencies. These issues are likely to result in downtime, security vulnerabilities, and inefficient resource utilization.
Integration testing with Dapr and Testcontainers
This blog post discusses setting up a Pytest integration test framework for a system using Dapr and Testcontainers. This framework was initially setup for a customer to suit their needs and it has been extracted into a sample project to provide a starting point.
Running RAG with ONNX Runtime GenAI for On-Prem Windows
Exploring how to efficiently run a RAG pipeline with structured language models (SLMs) and guardrails on Windows, achieving inference under 5 seconds with ONNX Runtime GenAI.
Data Validations with Great Expectations in MS Fabric
In this blog post, we will explore how to integrate GX within the Microsoft Fabric environment and use it to validate data programmatically.
Using OpenTelemetry for flexible observability
OpenTelemetry offers flexible observability in distributed systems, this post looks at how to integrate multiple data sources when connected and disconnected from the internet.
The Hidden Risks of Docker Build Time Arguments and How to Secure Your Secrets
This post discusses the security risks of using Docker build-time arguments to pass secrets, and suggests alternative methods for handling secrets securely.