@japkeerat

Building Production-Ready Agentic Systems: Lessons from Real Implementation

Aug 27, 2025 · 13 min read · Most articles about agentic systems show you the glossy marketing version. Multi-agent orchestration! Intelligent decision making! Autonomous task execution! The reality is more nuanced. After building a production agentic system that handles real El...

I finally built my first MCP Server. I expected magic, I got glorified API endpoints

Aug 16, 2025 · 8 min read · It is 2025. Everybody's releasing their MCP Server. I, on the other hand, was avoiding it like Neo avoided the red pill. But just like in The Matrix, you can only dodge inevitability for so long. And I built one. For BigQuery. Even though Google's MC...

Let's talk about Perplexity

Jan 2, 2025 · 6 min read · The Generative AI race has been coupled with a rise in usage of the term “Perplexity”. Google Trends suggests the same and most of the references in academic journals coming from the last 1 year. And no, this is not perplexity.ai. It is a metric bei...

1

What I learned about PyPi from maintaining an Open-Source Package

Dec 26, 2024 · 6 min read · Last month, I published a package on PyPi - TezzCrawler - a simple CLI tool to convert any website to LLM ready draft for building a RAG capabilities on any website. What spiraled next, was hundreds of hours of analysis on how PyPi works and what hap...

LSadra commented

Is the Model making right predictions? - Part 5 of 5 on Evaluation of Machine Learning Models

Dec 23, 2024 · 5 min read · Preparing your dataset, especially the test set, is a crucial step in building reliable and high-performing machine learning models. Proper dataset preparation ensures that your model is not only accurate but also generalizable to new, unseen data. L...