Saurabh Sarkar·2d agoHow Running an AI Company Changed the Way I Think About Engineering LeadershipWhat building an AI company taught me about uncertainty, evaluation, and leading teams that deliver real outcomes.
InThe ISO NavigatorbyManish Patel·Oct 23ISO 20022 Navigator is now also available as an MCP serverAI assistants explain ISO 20022 well, but exact codes, message versions and scheme mappings are easy to get wrong from memory.
InUnderdog EngineeringbyJames Wang·Oct 23Everyone ShipsSix months of letting non-eng build real software at Underdog
Kartik sarda·Oct 21Benchmarking Speculative Decoding in vLLM V1: From a 103ms Regression to +20% ThroughputSpeculative decoding promises a compelling trade: pair a small, fast draft model with a large target model, propose K tokens ahead, and…
Yang Liu·Oct 20Building My Creative Studio is teaching me about applied AII started Kiteva because I wanted to explore something fun: using AI agents and code to turn an idea into a video. I have spent years as a…A response icon1A response icon1
Paolo Rechia·Oct 18Building my own local voice agent — part 1How is this article structuredA response icon1A response icon1
sourav mansingh·Oct 16Your AI Agent Needs a Memory That Survives Crashes: A Practical Guide to TemporalWhat Temporal is, why it exists, and where it fits in real Applied AI systems
Kiell Tampubolon·Oct 13How to Build a Financial Filings Pipeline With OCR, Extraction and EmbeddingsQuick answer: Build a financial filings pipeline from structured sources first: pull XBRL facts, parse HTML tables for what the tags miss…
InThe ISO NavigatorbyManish Patel·Oct 5Reading camt.053 with an LLMI was surprised how much of the message a bank can omit.
