Series

Artificial Intelligence

Hands-on learnings, practical experiments, and engineering observations from my journey exploring Artificial Intelligence, RAG systems, LLM workflows, and real-world AI application design.

I Built a RAG Pipeline and Learned That Retrieval Is the Real Product.
Why building with LLMs taught me that the model is only one part of the system
May 8, 202626 min read24
What actually happens between your prompt and the final AI answer
When most people use an AI application, the experience feels almost magical. You type a prompt.You wait for a few seconds.And then a polished answer appears on the screen. From the outside, it feels s
May 8, 202624 min read7
Why Most AI Demos Fail in Production !
A working prompt is only the beginning. The real challenge starts when reliability, retrieval quality, latency, observability, and user trust enter the picture.
May 11, 202612 min read1
The Hidden Distributed System Inside Every AI Feature !!
Most AI failures in production are not just model issues. They are distributed systems failures expressed through an AI interface.
May 13, 202643 min read4
Designing Production-Ready AI Agents with MCP, FastAPI, and LangChain !!!
Why Most Enterprise AI Agents Fail Before They Reach Production A lot of AI demos look impressive in isolation. A chatbot answers questions, a tool-calling agent fetches data, or a retrieval pipeline
May 29, 202616 min read10
Why Local AI Stacks Get Hard Fast: Ollama, FastAPI, Vector Search, and the Operational Tradeoffs Nobody Mentions !!!
Introduction I used to think local AI architecture was one of those rare wins in engineering that looked simple on paper and actually stayed simple in practice. You run a model locally with Ollama, ex
Jun 7, 202622 min read8

Command Palette