ME

Software Engineer

shawaf.me

Architecture notes, delivery lessons, and career signals

A developer-first journal of shipped work

Everything I publish here mirrors the way I work: pragmatic, traceable, and ready for production. Articles are pulled directly from my Medium feed so you can read them here without leaving the portfolio.

Medium FeedJuly 24, 2026

GenAI-PART4: Beyond Words: How AI Learned to See, Draw, Hear — and Use Every Sense at Once

Vision Transformers, CLIP, diffusion models, AudioCraft, and Gemini — the complete story of AI’s new senses Type “a cat riding a unicorn in space, cartoon style” into an image generator and something remarkable happens twice: the machine understands your sente

#generative-ai#llm
Medium FeedJuly 21, 2026

GenAI-PART3: Scaling Smarter, Not Bigger: How Modern LLMs Got So Capable Without Bankrupting…

GenAI-PART3: Scaling Smarter, Not Bigger: How Modern LLMs Got So Capable Without Bankrupting Everyone Mixture of Experts, reasoning models, and million-token context windows — the three tricks behind today’s frontier AI For a few years, the recipe for a better

#generative-ai#llm
Medium FeedJuly 18, 2026

GenAI-PART2: The Life of a Foundation Model: From Raw Data to the AI in Your Pocket

How one model gets born, educated, specialized, aligned, and shrunk — the full lifecycle behind every AI product you use Every AI assistant you’ve ever used — ChatGPT, Claude, Gemini, the autocomplete in your IDE — is the end of a story. Behind it sits one eno

#llm#generative-ai#foundational-model#ai
Medium FeedJuly 9, 2026

GenAI-PART1: From Counting Words to ChatGPT: The 60-Year Road to LLMs in One Article

Every breakthrough in AI language exists because the previous idea hit a wall. Here’s the whole chain — in about 8 minutes. Open any product review section and you’ll find something like: OMG!!!!! BEST. PRODUCT. EVER?!?!?? 🔥 You understood that instantly. To

#ai#llm#generative-ai#nlp
Medium FeedJanuary 14, 2026

Automation Will Force You to Understand Your System — Whether You Want To or Not

Automation Will Force You to Understand Your System — Whether You Want To or Not No Real System Diagram Here, Just Sample Automation rarely enters a system when everything is clean and calm. In complex ML and backend platforms, it usually arrives when things h

Medium FeedDecember 2, 2025

JavaScript: Event Loop Orchestration

JavaScript is famously single-threaded, yet somehow remains fast, responsive, and capable of handling network requests, timers, user interactions, and animations all at once. It doesn’t block the UI while waiting for slow operations, and promise callbacks some

#javascript#web-development#asynchronous-programming#event-loop
Medium FeedMarch 23, 2025

Room Database: Migration Solutions

📌 Introduction Room Database is a go-to solution for handling local storage in Android apps, offering a seamless abstraction over SQLite. But what happens when your schema changes? Without proper migrations, users might lose data or experience crashes when up

#room#android#database#mobile-app-development
Medium FeedJanuary 26, 2025

Jetpack Compose Lifecycle: Around Composition Process

Hey there! If you’re diving into Jetpack Compose, Google’s modern UI toolkit for Android, like I did a few months ago ( we late to the party right :D ), I will help you understand a really important part of how compose lifecycle works. Don’t worry if it feels

#jetpack-compose#android#mobile-development
Medium FeedJanuary 22, 2025

Understanding @Immutable in Jetpack Compose: When and Why to Use It

When developing modern Android apps with Jetpack Compose, you’ll often hear about `@Immutable` and wonder what purpose it serves. If you’re unsure why it’s needed or confused about how it works, you’re not alone. This article explains `@Immutable`, its role in

#mobile-development#jetpack-compose#immutability#android