Data Collection, Machine Learning — But No AI Yet : Why today's frontier models are the most impressive recorders ever builtToday's large language models do not learn the way the word is sold to the public. They record and compress trillions of tokens of human writing — scraped, copied, and in documented court cases pirate12d ago · 41 min readramitdatta.com
How GDTM and CDTM Reshape Governance and Corporate EcosystemsThe Problem That Won't Go Away Every meaningful organization — a startup, a multinational, a city council, a country — runs on a flow of decisions taken under uncertainty across many time horizons by May 4 · 16 min readramitdatta.com
Why We Built DB Agents Instead of Using an Agent FrameworkA deep-dive into production-grade multi-agent architecture on Cloud — and why off-the-shelf frameworks couldn't get us there. The Problem with Agent Frameworks Every week, a new LLM agent framework aMar 16 · 24 min readramitdatta.com
The Complete Guide to SQL Vector Databases: From Theory to Production-Ready SQL ImplementationsWhat is a Vector Database? A vector database is a specialized database system designed to store, index, and query high-dimensional vectors (arrays of numbers) that represent complex data like text, images, audio, or sensor readings. Unlike traditiona...Jan 9 · 5 min readramitdatta.com
Vector Filesystem Search: Finally, Files That Find ThemselvesI built a vector search service for all platforms and filesystems called vecfs. It’s a little daemon that runs in the background and finally treats file discovery like a real systems problem—not something you bolt on after the fact. Instead of just l...Dec 15, 2025 · 4 min readramitdatta.com
