InRiafy StoriesbyJohn Mathew·Oct 29, 2025Make Failures Boring: Reliability at Trillion ScalePart 2 of A Trillion-Token Operator’s Playbook
InRiafy StoriesbyJohn Mathew·Oct 27, 2025A Trillion-Token Operator’s PlaybookLessons we learned from handling trillions of tokens in productionA response icon1A response icon1
InRiafy StoriesbyJohn Mathew·Oct 23, 2025Part 2 | The Performance Paradox — When “Smarter” Models Break What WorkedThe impact of AI Model retirement
InRiafy StoriesbyJohn Mathew·Oct 22, 2025Part 1 | The Hidden Price of “Better” — When Model Deprecation Tests Production FaithThe impact of AI Model retirement
InRiafy StoriesbyJohn Mathew·Sep 17, 2025I Bet Your GenAI Will Fail (Unless You Do These 4 Things)MIT’s Research says 95% Enterprise AI use cases get Zero ROIA response icon1A response icon1
InRiafy StoriesbyJohn Mathew·Mar 12, 2025The Hard Truth About AI Search ToolsWhat the Tow Center for Digital Journalism revealed about generative search accuracy
InRiafy StoriesbyJohn Mathew·Mar 5, 2025The Hard Truth About AI ReasoningWhat Riafy shared at Google’s Gen AI Live 2025
InRiafy StoriesbyJohn Mathew·Feb 17, 2025DeepSeek: The Illusion of Cost Savings and the Hidden Price of AI SafetyAI is supposed to make things easier. Smarter. Safer. But what happens when an AI that promises cost savings turns out to be a liability?
InRiafy StoriesbyJohn Mathew·Nov 6, 2024When Silence Speaks Louder Than Words: The R10 Deep StoryIn the race for AI advancement, we’ve been asking the wrong question. Instead of “How much can an AI answer?”, maybe we should have been…
InRiafy StoriesbyJohn Mathew·Nov 6, 2024‘Trust Me, I’m an AI’: How SimpleQA Exposed the Gap Between Confidence and TruthAn OpenAI Benchmark that has caught LLMs in a `Web of Lies`
