top of page


The Design Space of an LLM System
Where this starts When we design a system that uses agents as one of its components, we connect it to things the way we always have: an external service, a database, an API. Each of those comes with a protocol, a fixed shape for what you send and what you get back. The connection to an LLM has none of that. There is no fixed request format, no fixed response format, no fixed type for the input or the output. That makes the space of possible designs very wide, and very easy to
NOURA ALSHAREEF
Oct 3113 min read


Neural Network Search - NAS
For years, neural network architectures were designed by hand. The space of possible designs is enormous: how many layers, how wide each one is, how many branches, which kernel sizes, what input resolution. Turning all of those knobs manually, and getting them right together, is famously difficult. Neural architecture search (NAS) hands the tuning to an algorithm, which explores those knobs under whatever accuracy and efficiency constraints we set. That saves a great deal of
NOURA ALSHAREEF
Oct 2624 min read


The building blocks NAS chooses from
Oct 23, 2026 · @Nora 1. Quick recap: what a network is made of Say we have data for 100,000 patients, and for every patient five measurements. Those five numbers are the inputs x₀…x₄ of one data point. We do not feed patients one by one; we take them in batches, maybe 100 at a time, so one batch is a table with 100 rows and 5 columns. In the picture below we use 2 at a time so it stays readable: patient A in front, patient B behind. The data decide the values in the circles.
NOURA ALSHAREEF
Oct 2415 min read


Quantization, Part 2
Oct 22, 2026 · @Nora 1. Where we stopped Last time we laid a ruler over the weights: evenly spaced marks, every weight snaps to the nearest one, and two numbers (scale and zero point) bring it back. That is linear quantization, and it is what lets a layer run on integers. But the ruler left three questions open, and they decide whether the quantized model keeps its accuracy: One ruler or many? A single ruler for a whole layer, or one per channel, or one per small group of wei
NOURA ALSHAREEF
Oct 2313 min read


NN Quantization
1. Why fewer bits? Pruning made the network smaller by deleting weights. Quantization keeps every weight but stores each one in fewer bits. The two stack: a pruned network can also be quantized. Why it pays off: Memory. A weight in 32-bit float takes 4 bytes. The same weight in 8-bit integer takes 1 byte: the model is 4× smaller and moves 4× less data from memory, which in Why Is My Neural Network Slow? we saw is where most of the time and energy go. Energy. An 8-bit integer
NOURA ALSHAREEF
Oct 2310 min read


How Much to Prune, and How to Recover
1. Where we stopped last time In Part 1 we saw what pruning is: delete the weights the network does not need. In Part 2 we answered two of the four questions: in what shape do we delete (granularity), and which weights go (criterion). Two questions are left, and they are the ones you meet first when you actually prune a model: How much do we prune in each layer? Pruning every layer by the same amount is simple, but it is not the best. Some layers can lose 90% of their weights
NOURA ALSHAREEF
Oct 239 min read


A Deep Dive into Pruning Patterns
A simple guide to choosing the pattern and the weights to prune In this article Where we stopped last time Granularity: fine vs coarse Five patterns in a convolution layer 3.1 Fine-grained: any weight, anywhere 3.2 Pattern-based: zeros in allowed shapes 3.3 Vector-level: a whole row of a kernel 3.4 Kernel-level: a whole 3×3 kernel 3.5 Channel-level: a whole filter Criterion: which weights go? 4.1 Magnitude 4.2 Wanda: weights × inputs Test yourself: one panel, five patterns 1.
NOURA ALSHAREEF
Oct 1511 min read


NN Pruning
A simple guide to removing the weights a neural network does not need In the last post we used the roofline to find what slows a network down. One of the fixes was a single line: pruning means removing weights that don't matter. Fewer weights, fewer MACs. This post explains that line. It answers two questions: What is pruning? How do we write pruning as a math problem? Why remove weights at all? Every weight costs us twice. It takes bytes in memory, and it takes one MAC every
NOURA ALSHAREEF
Oct 138 min read


Why Is My Neural Network Slow?
A simple guide to matrix multiplication, hardware, and the roofline 1. It is all matrix multiplication A neural network does a lot of math. Most of that math is one thing: matrix multiplication. A layer in a network takes an input and multiplies it by a table of numbers called weights. That is a matrix multiplication. Most of the time a network runs, it is doing this. So if we want fast AI, we need fast matrix multiplication. الشبكة العصبية تقضي معظم وقتها في ضرب المصفوفات، ل
NOURA ALSHAREEF
Sep 2512 min read
Intro to LLMs
Andrej Karpathy's "Intro to Large Language Models" came out in 2023, and I still go back to it. The details have aged, but the story hasn't. It gives you the full picture of what an LLM is and how the ideas connect, from "a model is just two files" to pretraining, finetuning, and RLHF. Every time I rewatched it, the same questions came up: what actually happens inside the network? How does text become numbers? What is attention doing? So I turned the first part of the talk in
NOURA ALSHAREEF
Sep 241 min read


An Interactive Guide to CNNs
I always wanted one place with everything about CNNs: the details, the visualizations, and the chance to play with the numbers. So I built it: an interactive reference where every step, from pixels to prediction, is computed live and explained in plain words. I hope you enjoy it. You can also experiment with a larger width at: https://nora-alshareef.github.io/cnn-interactive-guide/
NOURA ALSHAREEF
Sep 191 min read


Money-for-startups: 📋 SAFE Template
A quick safe template
NOURA ALSHAREEF
Apr 252 min read


Money for Startups: The SAFE
At the early stages — pre-seed and seed — you may have incorporated the company, you probably have a co-founder, and you almost certainly have very little revenue. What you have is conviction. And that's mostly what you're selling. The people who show up to invest at this stage know that. Friends and family. Angel investors who back people more than products. Accelerators like Y Combinator. Sometimes your own savings. They're betting on you — because there isn't enough of a b
NOURA ALSHAREEF
Apr 244 min read


Money for startups : The Funding Stages - Pre-Seed, Seed, and Series A
In the first two articles, we covered the big picture — how money moves into a company, what you're giving away when you take it, and the documents that govern that exchange. We talked briefly about SAFEs, term sheets, valuation, dilution, and option pools — we'll cover each in detail in the articles ahead. But we kept saying things like "at your first round" or "when a VC comes in" without stopping to explain: when does any of this actually happen? What does the sequence loo
NOURA ALSHAREEF
Apr 242 min read


Money for startups: The Option Pool Trap Nobody Warns You About
In Nine Steps Every Founder Must Take Before Seeking an Investor, we listed the option pool as step nine — the final step before bringing in outside capital. By that stage, you have already: ✅ Agreed on the founder equity split and vesting ✅ Incorporated the company ✅ Issued founder shares with vesting ✅ Filed your 83(b) election (within 30 days of the grant date) ✅ Signed IP assignment agreements The foundation is set. Ownership is clean. The structure is legal. Then comes t
NOURA ALSHAREEF
Apr 246 min read


Money for Startups: Vesting
In the previous article, we built the cap table — the document that records who owns what in your company. We saw how founders agree on an equity split, authorize shares, and issue restricted stock. We mentioned vesting as something attached to those shares, but we moved past it quickly. Now we are going to slow down and look at it properly. Vesting is one of those concepts that sounds punishing the first time you hear it. You think you own 50% of the company. You do not — no
NOURA ALSHAREEF
Apr 246 min read


Money for startups : Cap table
In the previous article, Nine Steps Every Founder Must Take Before Seeking an Investor, we walked through everything that needs to happen before you ever sit across from a VC. Once you’ve built the product and identified your customer, the work becomes structural: hire a lawyer, agree on the founder equity split and vesting, incorporate, issue founder shares with vesting, file the 83(b) (if you’re in the U.S.), sign IP assignment agreements, and create an option pool. Now we’
NOURA ALSHAREEF
Apr 234 min read


Before the Money: Nine Steps Every Founder Must Take Before Seeking an Investor
In the first two articles of this series : How Startups Get Funded and How the Game Actually Works , we covered the big picture — how money moves into a company, what you give away when you take it, and the documents that govern that exchange. We touched briefly on SAFEs, term sheets, valuation, dilution, and option pools, all of which we will cover in detail in upcoming articles. In the third article, we built a glossary you can return to whenever a term feels unfamiliar. No
NOURA ALSHAREEF
Apr 234 min read


The Startup Glossary: Words You'll Keep Hearing (and What They Actually Mean)
If you've been reading this series — or talking to investors, lawyers, or other founders — you've probably hit a wall of terminology that no one stops to explain. Everyone just assumes you know what a liquidation preference is. They assume "pro-rata" makes sense to you. They drop "SAFE" into a sentence and move on. This article is the one we'll link back to every time we use a term you might not be familiar with. Bookmark it. Come back to it. It'll make more sense every time
NOURA ALSHAREEF
Apr 224 min read


Money for Startups: How the Game Actually Works
Most founders think venture capital is about money. It's not. It's about understanding what you're trading — ownership, control, and sometimes the CEO seat — before you walk into that room. This is the map.
NOURA ALSHAREEF
Apr 208 min read

STAY IN THE KNOW
bottom of page