The Engineering Notebook — page 27 of 32

Notes on building with AI, agents & the modern stack.

Deep dives on AI/ML, RAG systems, agent engineering, and senior-engineer architecture decisions — a new post every week.

Police station entrance at night with lights on Cybersecurity

Microsoft's FedRAMP Failure: A Technical Breakdown of How a 'Pile of Shit' Cloud Got Approved [2026]

A former Microsoft engineer called Azure Government a 'pile of shit.' Here's which FedRAMP controls allegedly failed, why the system didn't catch it, and what it means for every engineer selling to the public sector.

jemalloc, tcmalloc, performance-engineering, p99-latency, memory-allocator Technology

jemalloc vs malloc vs tcmalloc: Why Your Server's Default Allocator Is Killing P99 Latency [2026 Guide]

Your server's default memory allocator is almost certainly the wrong choice for concurrent workloads. Here's how jemalloc and tcmalloc crush P99 latency without changing a single line of code.

batteries, lenovo, laptops, fast-charging, storedot Technology

Silicon-Anode Batteries: What Lenovo's StoreDot Bet Means for 20-Minute Laptop Charging [2026]

Lenovo invested in StoreDot's silicon-anode battery tech that could charge a laptop in under 20 minutes. Here's why this matters and when you'll actually see it.

a black house with a green light in the window Technology

Self-Hosted Voice Assistant With Home Assistant: The Complete 2026 Guide to Ditching Alexa

Amazon and Google are stuffing ads and unwanted AI into your voice assistants. Here's how to build a fully private, locally-hosted alternative with Home Assistant that actually works.

amd-rocm, cuda, local-ai, local-llm, gpu, open-source-ai, ollama, llama-cpp, ai-hardware, rx-7900-xtx, hip Technology

AMD ROCm vs CUDA for Local AI [2026 Compared]

After months of running LLMs on AMD hardware, here's the brutally honest comparison of ROCm vs CUDA that no marketing page will give you — covering real performance, setup friction, and which GPUs actually work.

Computer screen displaying lines of code Cybersecurity

Glassworm Is Back: The Invisible Unicode Attack Hiding in Your Code

Your code reviewer can't catch what they can't see. The Glassworm invisible Unicode attack is resurging, and most repos have zero defenses against it.

Abstract purple lines on a black background AI and Machine Learning

How to Build an AI Agent With Python in 2026: Stop Building Solo Agents, Start Building Teams

Single-agent LLM wrappers are already obsolete. In 2026, the real power move is orchestrating teams of specialized AI agents. Here's the production-ready blueprint.

A mechanical keyboard sits on a desk mat. Cybersecurity

Kernel-Level Anti-Cheat: How Ring 0 Drivers Actually Work and Why Gamers Are Right to Be Nervous

Modern anti-cheat software runs at the deepest level of your operating system. Here's exactly how it works, why developers say they need it, and why the security tradeoffs are real.

chrome, cybersecurity, zero-day, google, browser-security Technology

Chrome Zero-Day Exploits: 4 Vulnerabilities Exposed in May 2024's Most Aggressive Attack Wave

Four Chrome zero-days exploited in the wild within two weeks. Here's what each vulnerability actually does, who was targeted, and what this signals about browser security.

semiconductors, supply-chain, helium, manufacturing, geopolitics Technology

Helium Shortage and Semiconductor Supply Chain: The Hidden Crisis Nobody's Talking About [2026]

Everyone's worried about silicon wafers and ASML machines. The real bottleneck in chip manufacturing is a noble gas that literally floats away when you're not looking.

Abstract flowing lines of purple and white on black Breaking Tech News

Meta's Default Encryption for Messenger: Why Governments Are Furious and What It Actually Means

Meta finally flipped the switch on default end-to-end encryption for Messenger. Billions of messages are now unreadable by anyone but sender and recipient. Law enforcement is furious.

a close up of a microphone on a table AI and Machine Learning

Amazon Wants to Charge for Alexa. The Real Risk Isn't the Price. It's the Personality.

Amazon's splitting Alexa into free and paid tiers. This isn't just monetization. It's a bet that personality-driven AI justifies a subscription. That's harder than it sounds.