Writing
Production engineering — the decisions, trade-offs, and systems that don't show up in tutorials.
Software Debugging Techniques: Which One Do You Reach For at 2 AM?
Most engineers have one go-to method. The best engineers have a playbook of software debugging...
Our Benchmark Reported 0.96 Recall. It Was Measuring Nothing.
An overnight benchmark run produced a 0.9653 and a 0.0000 for the same dataset. Both numbers were wrong, for different reasons. A postmortem on four measurement bugs — and why every one of them reported success.
What If AI Stopped Talking? Inside Jev, System One Models, and RLCD
LLMs are great at generating text. But many software systems do not need generated text.They need decisions.For example:Is this transaction suspicious?Which team should handle this ticket?Is this RAG …
Your Agent's Memory Shouldn't Answer Questions It Wasn't Asked
Every agent-memory demo stores a question and retrieves it back. Here's the query that breaks most of them — and why I think memory needs a decision layer, not just a retriever.
I Built an Incident-Triage Agent That Resumes From Checkpoints
Most LLM agents lose everything when they crash. AgentCore checkpoints every pipeline step to Postgres and writes resolved incidents into episodic memory — here's the architecture, and what it actually learns.
I Built a Developer Portfolio With AI and GitHub Pages — For $0
No hosting bill. No design expertise. Just a few hours, GitHub, Next.js, and AI.Most developers know they should have a portfolio.So they follow the same tutorial:Pick a templateChange the nameAdd thr…
Build a Free Developer Portfolio with AI and GitHub Pages — In an Afternoon
No hosting bill. No design skills required. A step-by-step guide to shipping a real portfolio link you can hand to an interviewer today.
I Got Tired of Benchmarks That Test Models Instead of Agent Frameworks, So I Built My Own
Every benchmark tells you which model is best. None of them tell you which agent framework is best for your actual daily work. So I built one that does.
Building an A2A Client: What the Protocol Gives You — and What You Still Have to Build
Agent interoperability is only one layer of a multi-agent system. Discovery, routing, authentication, execution, and failure handling are still your application’s job.AI agents are becoming increasing…
What Nobody Teaches You After You Learn to Code
The gap between tutorial engineer and production engineer is real, wide, and almost entirely undocumented. That's what TheProdSDE is here for.
LLMs Under the Hood: KV Cache, VRAM, Tokens & Inference Explained
From RNNs to PagedAttention — everything that actually matters, explained with intuition first.Understand how modern LLMs work under the hood — transformers, tokenization, VRAM, KV Cache, quantization…
A Production Engineer’s Guide to Choosing the Right Prompting Strategy
Most prompt engineering articles explain techniques: Zero‑shot, Few‑shot, Chain‑of‑Thought, ReAct, Tree‑of‑Thought.Production engineers need something different: a framework that tells them which prom…
Why NVIDIA NIM’s Free API Feels Slow (And What Your Upgrade Path Actually Looks Like)
If you’ve used NVIDIA’s free API through build.nvidia.com — part of their NIM (Inference Microservices) platform — you’ve likely hit frustrating latency. Before you “upgrade,” it’s worth understanding…
Cutting My Subscription Cost Without Impact on Workflow — Here’s the Exact System I Use
Most developers are haemorrhaging money on AI tools. This one mindset shift changed everything.TL;DR: Routing your prompts to the right AI model — premium for reasoning, free for execution — can cut y…
I ran two AI builders in parallel at a solo hackathon — here's what the 429 errors revealed
At a 3-hour Google for Developers hackathon, I ran two AI builders simultaneously on the same...
One Prompt vs Six: I Ran Two AI Builders in Parallel at a Hackathon
A real build log from PromptWars (Google for Developers): autonomous agent vs spec-first workflow, how each one handled rate limits differently, and the exact prompts I used — word for word.SEO descri…
Senior AI Interviews Don’t Test What You Know. They Test What Breaks at 2am.
15 questions that separate engineers who’ve shipped production AI from those who’ve only read about it — and the answers that actually hold up when the latency doubles overnight and nobody knows why.M…
Zero Trust Security in Production: Identity, OPA, Vault, mTLS & Audit Logging — A Complete Reference
This is **Part 5* of the Building a Zero-Trust Security Architecture series. Parts 1–4 covered...
Your App Has 6 Security Layers. You’re Probably Collapsing Them Into 1.
Authentication, authorization, secrets, encryption, audit, and network control are six different jobs — and collapsing them is the #1 architecture mistake I keep seeing in production.Series: Part 1 · …
Azure Key Vault vs AWS Secrets Manager: Cutting Through the Cloud Marketing
"A hands-on breakdown of Azure Key Vault, AWS Secrets Manager, IRSA, Azure Managed Identity,...
I Built a Finance App at a Hackathon, Abandoned It for Months, Then GitHub Copilot Helped Me Finally Ship It
This is a submission for the GitHub Finish-Up-A-Thon Challenge What I Built "Where did...
Your JWT Is Lying to You — The Authorization Problem Nobody Solves Correctly
A valid token proves who you are. It says almost nothing about what you're actually allowed to...
Vibe-Coding Works. That's Exactly Why It Will Destroy Your Codebase at Scale.
The productivity gains are real. The compound debt is realer. And almost no team is measuring the...
MCP for Backend Engineers: When to Use It (and When to Skip It)
"If you only have one consumer, MCP is a liability." Subtitle: Real code, real cost, and an honest...
Worth Each minute read if you work on RAG
Why Most RAG Systems Fail in Production (And How to Design One That Actually Works) ...
Agentic AI Fails in Production for Simple Reasons — What MLDS 2026 Taught Me
TL;DR: Most agentic AI failures in production are not caused by weak models, but by stale data, poor...
Why Most RAG Systems Fail in Production (And How to Design One That Actually Works)
A practical, system design–focused breakdown of why RAG systems degrade after launch—and what...
Most AI Agent Frameworks Are Overkill — Here's How to Choose the Right One in 30 Seconds
A senior engineer's field-tested breakdown of LangGraph, AutoGen, CrewAI, Microsoft Agent Framework,...
[Boost]
Stop Guessing Your LLM Replacement TheProdSDE ・ Mar 14 ...
Stop Guessing Your LLM Replacement
A Practical Guide to Migrating GPT Apps Across Azure, AWS, and GCP TL;DR — Most LLM...
Worth your 10min of time if deciding which framework suits your use case better
LangGraph vs Semantic Kernel: Python AI Agents in 2026 ...
[Boost]
7 AI Prompts Senior Engineers Use to Build Production Code...
7 AI Prompts Senior Engineers Use to Build Production Code Faster
AI coding tools are powerful — but most developers still use them like autocomplete. The difference...
Stop Treating AI Like Autocomplete — Design an AI-First Developer Workflow Instead
AI coding tools are everywhere — but most teams still use them like autocomplete. The result? Faster...
LangGraph vs Semantic Kernel: Python AI Agents in 2026
Why This Comparison Matters Right Now Here's the honest reality of building Python AI...
Azure App Service, Container Apps, and AKS compared: Real costs in INR, architecture constraints, and when to upgrade for production micro-services
If you’ve ever tried to deploy a full stack (Node backend, MCP server, ETL jobs, Angular SPA, and...