Skip to content

Writing

Long-form ideas on enterprise AI, GenAI execution, and leadership.

Medium posts

2026-03-09

Speed, Precision, and Late Interaction: A Deep Dive into Bi-Encoders, Cross-Encoders, and ColBERT

How to optimize RAG systems by balancing retrieval speed with search precision using a "funnel" architecture. It details the unique roles of Bi-Encoders for rapid initial search, Cross-Encoders for high-accuracy reranking, and ColBERT as a middle-ground solution for efficient late interaction.

Read on Medium ↗

2026-02-18

RAG Chunking Strategies: A Complete Guide for AI Architects

Practical approaches to chunking and retrieval for RAG pipelines: tradeoffs, chunk size guidance, and deployment considerations.

Read on Medium ↗

2025-02-03

Demystifying the Transformer

A plain-English walkthrough of Transformer encoder internals and why attention mechanisms matter for sequence modeling.

Read on Medium ↗

2024-05-31

Securing Azure Databricks Communication

Technical walkthrough for hardening the connectivity between the control plane, data plane, and end users. The guide focuses on eliminating public exposure by implementing Secure Cluster Connectivity (SCC) and utilizing Private Endpoints to ensure all workspace interactions remain within a private network.

Read on Medium ↗

2023-04-25

Azure Storage Encryption and Azure Disk Encryption

This article clarifies the often-confused distinction between Server-Side Encryption (SSE) at the storage level and Azure Disk Encryption (ADE) at the OS level. The article provides a comprehensive guide on managing encryption keys and features a decision flowchart to help architects choose the right model for securing data at rest, in transit, and in use.

Read on Medium ↗

2023-01-05

Availability Sets in Azure

How to safeguard virtual machines against hardware failures and maintenance outages by distributing workloads across Fault Domains and Update Domains. The article also introduces Proximity Placement Groups as a specialized solution for achieving the sub-millisecond latency required by high-performance applications.

Read on Medium ↗

2022-12-15

Azure App Registration, Enterprise Apps, and Service Principals

The relationship between global App Registrations and local Service Principals, explaining how they function together to manage identity and access. The article provides a clear distinction between how these objects behave in single-tenant versus multi-tenant scenarios, specifically highlighting how permissions and consent are handled across different organizations.

Read on Medium ↗

2022-11-16

Azure Kubernetes Networking Services and Ingress Controllers

Explains how to bridge the gap between ephemeral pods and stable external access using various Service types and Layer 7 Load Balancers. The article highlights the architectural differences between Nginx and Azure Application Gateway (AGIC), specifically how they handle traffic routing and interact with the Azure control plane.

Read on Medium ↗

2022-11-03

Overview of Azure Kubernetes Services Networking Models

Provides a comprehensive overview of the networking models available in Azure Kubernetes Services (AKS), including the differences between Azure CNI and Kubenet, and how these models impact pod communication, network security, and integration with other Azure services.

Read on Medium ↗