Saurabh Kumar https://saurabhkumar8112.github.io/ Technical articles by Saurabh Kumar on AI agents, model evaluation, reliability, and cost, with practical guides, diagrams, and reproducible code. en Everything Can Be Automated with AI. Why Bother Then? https://saurabhkumar8112.github.io/articles/everything-can-be-automated/ https://saurabhkumar8112.github.io/articles/everything-can-be-automated/ AI multiplies a system. Without one, it multiplies variance. An interactive essay on reels, assembly lines and why judgment is the new bottleneck. Mon, 19 Oct 2026 12:00:00 +0900 When AI Does the Work, What Will Infosys and TCS Sell? https://saurabhkumar8112.github.io/articles/future-of-it-services/ https://saurabhkumar8112.github.io/articles/future-of-it-services/ What Infosys and TCS earnings calls reveal about AI, outcome pricing and productivity, and how agent evaluations can make delivery commitments credible. Wed, 23 Sep 2026 21:59:56 +0900 When should you use JEV? A practical guide https://saurabhkumar8112.github.io/articles/when-should-you-use-jev/ https://saurabhkumar8112.github.io/articles/when-should-you-use-jev/ When should you use TypeSafe's Jev? A practical guide to AI agents, search and coding, covering accuracy, error cascades, confidence and workflow costs. Tue, 22 Sep 2026 00:20:44 +0900