I thought my PDF parser was done — then I ran it on 80 real resumes
Building RAG pipelines, the step that kept letting me down was the very first one: turning a PDF into...
Tutorials
The two scariest parts of autonomous agents — runaway loops and exposed APIs
Everyone I talk to about autonomous agents has the same two fears, and they're both justified: The...
Your embedding axes can move while cosine neighbours stay put
An embedding tour shows how an orthogonal basis switch changes the view while every top-five cosine neighbour stays fixed.
DDD & More: Is Software Design Dead in the Age of Automatons?
Grab your favorite hot sauce, because the tech world is serving up some seriously fiery hot takes on...
Building Maestro AI: Routing LLM Calls So Your Agent Doesn't Burn Sonnet on Summaries
How I built a harness-agnostic model router for Cursor and Claude Code and what broke along the...
Turn any website into clean Markdown for your RAG pipeline (with code)
If you've built anything with retrieval-augmented generation (RAG), you know the unglamorous truth:...
I Got Obsessed With Killing Scrum. So I Built a World Out of 12 AI Agents.
This is a submission for Weekend Challenge: Passion Edition What I Built I built...
Claude Code's Auto Mode Now Default on Major Cloud Platforms
Anthropic's latest release makes autonomous coding the default on major cloud platforms, shifting the governance landscape for enterprise teams.
Your LLM Cannot Tell When It Is Wrong, Build for That
Every LLM hallucinates, and it is not a bug the next model release will fix. Next-token prediction...
Day 07: The Browser Environment & Git
🎯 Learning Objectives By the end of this article, you will understand: How to dissect a...
Stop Guessing: How I Pick AI API Architecture at Every Scale
Stop Guessing: How I Pick AI API Architecture at Every Scale I've been on both sides of this. Two...
Blocking AI crawlers earns you nothing. Here's how to price them instead
Disallow: GPTBot is a wall. Walls don't pay rent, and the crawlers that matter most either ignore...
7 things I learned trying to stop LLM API bills from silently exploding
My first real LLM bill surprise wasn't dramatic. No infinite loop, no viral spike. A retry policy I'd...
Why the codebase is your biggest context bill, and why output compression doesn't fix it
When people think about reducing AI coding costs, they usually think about the model's output: fewer...
Stop sending net-30 reminders: predicting when a client *actually* pays
Every freelancer knows the invoice that goes into the void. You did the work. You sent the PDF. "Net...
Test Mobile AI Input by Capability, Not Device Category
A capability matrix covers keyboards, styluses, foldables, voice, and interrupted networks better than phone-versus-tablet assumptions.
Describe your website once, make it usable by any AI
If you have tried to make a website work with AI assistants recently, you have run into the...
[GCP Billing & Vertex AI] Solving Gemini Cost Allocation in a Single Project: Vertex AI Dynamic Billing Labels in Action
Pain Point: How to Accurately Allocate Gemini API Costs Within the Same Project? When...
Hybrid Local + Cloud LLMs in 2026: When to Use Ollama and When to Pay for Fable
I run a local model and I pay for cloud models, and the most common question I get is "which one...
My Experience at Code on JVM Chennai: From DSA in Production to Building AI Products
Recently, I attended the Code on JVM Chennai meetup, and it was an incredible learning experience. As...
InsightsTrack + Pulse: I taught Claude Desktop to read my web analytics (via MCP)
This is a submission for Weekend Challenge: Passion Edition What I Built InsightsTrack is...
Your AGENTS.md Needs an Executable Contract, Not More Advice
Keep repository instructions honest by testing required paths, approved commands, and review dates in CI with a small shell-free Node.js checker.
I built a personal "Agentic OS" that runs my DBA work — with approval gates, audit trails, and a $0.01 morning brief
A local-first orchestration layer for AI agents, built by an Oracle DBA who doesn't trust anything that touches a database. Kernel, permission tiers, MCP tools, live dashboard.
5 Lessons From Letting My AI Coding Agent Open Its Own Pull Requests
TL;DR I let my autonomous coding agent open real pull requests without a human in the...