Two real AI failures, two domain safeguard designs, and the responsible-AI thinking that connects them. My write-up from a mini project on agent risk, ethics, and governance.
Our first build mirrored the org chart. It was the wrong shape. Here's how five recognised agentic design patterns, scored against the same criteria, led to a hybrid recommendation — and what changes in v3.
How I built LumenAI — a local-first, native macOS RSS reader with full-text search, offline reading, and pluggable AI summaries — from an empty folder to a signed DMG in about an hour, with Claude writing every line.
What ChatGPT and Claude actually are under the hood — a plain-English explainer of next-token prediction, softmax, attention, and why hallucinations are inevitable. Beginner to intermediate, with interactive animations.
The second chapter of Aether: how a local-first team of IT architecture agents grew from a clean idea into a 13-agent, web-first, self-escalating system — and every bug that shaped it along the way.
A visual breakdown of three RAG architectures — when each one wins, where it breaks down, and how binary quantization can shrink the vector index by 32× without changing the architecture you picked.
A six-minute executive briefing on the architectural and commercial shift underway in Indian infrastructure outsourcing. Where the industry sits on the automation maturity curve, why the incentive conflict is harder than the technology, and what one fictional 80,000-endpoint bank account looks like over a 24-month rollout. Companion to the long-form analysis.
Uber has moved Model Context Protocol from a pilot into a standardised internal platform that, by their own published figures, supports tens of thousands of weekly AI tasks across their internal service estate. This piece dissects what they appear to have built, what it changed, and — at greater length — what the same pattern could plausibly enable for Indian IT outsourcing in the infrastructure space. It separates what Uber publicly described from my own extrapolation, places the architecture in a six-stage maturity model, addresses the political and commercial incentive conflicts that make the rollout much harder than the diagrams suggest, and walks through one fictional 80,000-endpoint bank to make the economics concrete.
A domain-specific Retrieval-Augmented Generation assistant built with LangChain, OpenAI embeddings and FAISS that answers questions about the GitHub REST API strictly from indexed public documentation. Week 15 graded mini-project of the IITM Pravartak Professional Certificate Programme in Agentic AI and Applications.
