Thirumala kiran·Jun 26The End of the Coordination TaxOrganisations look the way they do because coordinating humans at scale is expensive. Agentic AI is about to make coordination cheap. The…
Thirumala kiran·Jun 14Everyone’s talking about GPUs. Nobody’s talking about the ground they run on.When OpenAI launches a new model, the internet erupts. When NVIDIA announces a new GPU, analysts scramble. But somewhere in Arizona right…
Thirumala kiran·May 24The Practitioner’s Guide to AI Evaluation ToolsEvaluating AI systems in production is not a benchmarking problem. It is an instrumentation problem. The question is not “how good is this…
Thirumala kiran·May 24Who governs the agent? A platform builder’s guide to AI governance across the stackAgentic AI doesn’t just raise new product questions. It raises new accountability questions — at every layer of the system you’re building…
Thirumala kiran·May 12Why OpenMetadata Is the Missing Layer in Your Agentic AI StackThe catalog isn’t just for data engineers anymore
Thirumala kiran·Apr 1Using GitHub Copilot Hooks at ScaleMost teams start with GitHub Copilot as an individual productivity boost. One engineer uses it for scaffolding, another for refactoring…
Thirumala kiran·Mar 18Throughput in LLMs: The Hidden Metric That Decides Whether Your System SurvivesMost discussions about large language models (LLMs) revolve around intelligence — reasoning ability, accuracy, hallucinations, benchmarks.
Thirumala kiran·Mar 4The 10 Hard Problems Every AI-Native Company Will FaceEveryone is building on top of LLMs right now.
Thirumala kiran·Feb 21What Are You Actually Asking Your AI Agents to Do?The Real Reason Most AI Initiatives Fail
Thirumala kiran·Jan 16Agent Autonomy Is a Spectrum: A Practical Maturity Model (L1 → L5)Over the past year, agentic AI has evolved from a niche research topic to a cornerstone of enterprise strategy. We’re bombarded with…