CloudYukthi

Cloud, with logic.

Practical writing on cloud architecture, AI infrastructure, multi-cloud, delivery, and technology strategy. The focus is on the reasoning behind decisions that hold up in production.

Explore the blog About CloudYukthi

Two perspectives

Explore CloudYukthi

Technical depth and leadership practice, connected by the same habit: reason about the constraint before choosing the solution.

Technical

Cloud architecture, multi-cloud, AI and GenAI, design reviews, migrations, cost optimisation, and production engineering.

Read technical posts

Leadership & Strategy

Cloud delivery, estimation, platform strategy, stakeholder alignment, and leading technology decisions under uncertainty.

Read leadership posts

From the blog

Latest writing

Practical perspectives on cloud, AI, architecture, delivery, and strategy.

View all posts

A pinned model being swapped for its successor behind a thin boundary, over the CloudYukthi gradient

When Your LLM Gets Retired: Keeping Model Choice Replaceable

Foundation model providers and cloud platforms retire models on aggressive schedules. Treating model selection as replaceable infrastructure across AWS, Azure, Google Cloud, OpenAI, and Anthropic turns an emergency fire drill into a routine config change.

generative-aiai-infrastructureai-model-governance

AdvancedNov 4, 2026Technical

About the author

Written by Pushparaju Thangavel

A technical leader and solutions architect writing about cloud, AI, architecture, and delivery from a practitioner perspective.

Read my story LinkedIn