Karim Shehadeh's Blog

I’ve been thinking lately about why some of the best ideas I’ve seen came from two or three people sitting around a whiteboard, while some of the most frustrating decisions came from rooms full of intelligent, experienced people.

At a certain scale, every successful SaaS company has to answer a quiet but important question: are customers staying because the product keeps getting better, or because leaving has become too painful?

The conversation around AI and productivity often feels oddly unsatisfying because both sides can point to evidence that feels true.

One thing I have noticed in large organizations is how easy it is to confuse motion with progress.

Most AI assistants feel impressive right up until you try to use them as part of your real daily workflow.

With the cost pressure of AI becoming a bigger news topic, it feels like a good time to talk about its overuse.

I’ve started to become pretty skeptical of heavily scripted approaches to AI-assisted software development.

The new cache model in Next.js is much more explicit than the older ISR/fetch-cache world.

One of the most common mistakes organizations make when approaching AI is starting from a place that is too broad to be actionable.

There's a notion that the To watch a machine produce reasonable answers to questions of all kinds as quickly as a slot machine spits out winnings is a disquieting thing.

AI has dramatically accelerated how quickly we can write software.

A product manager walks into a planning meeting with a feature idea.

We remake ourselves over our lifetime, usually as a result driven by personal change.

This week focused on major UX upgrades, better localization support, stronger profile/portfolio tools, and improvements across exhibitions and media handling.

One of the weirdest things about using AI to write software isn’t the speed, the convenience, or the “oh wow it actually worked” moments.

I spent most of yesterday building a pretty simple website called https://prompts.iwonderdesigns.com The idea came from my wife.

A maintenance-mode guide for days when nothing feels worth doing Some days don’t feel bad in a dramatic way.

For people who know me, it might be surprising to hear how much I love people 🙂.

I’ve been working on a tool meant to help me as an engineering manager, but somewhere along the way I realized it could be helpful to others too.

Deciding what type of technical solution to use is always shaped by three main constraints: the limits of technology, the capabilities of the team, and the long-term needs of the organization.

Like many engineering managers, my responsibilities shift from week to week.

I was recently thinking about why we get such different reports about the benefits of AI in software development.

I recently listened to One of his suggestions that resonated with me—and this is my interpretation—is using self-imposed micro time constraints for work.

Even though Trump has won the election, I believe we are about to witness an incredible groundswell of support for progressive policies.

Something I want to do on most of my versioned projects is handle the versioning and committing as part of the merge/push to the main branch.

The ability to juggle multiple tasks simultaneously is often celebrated as a hallmark of productivity.

I was inspired to talk a little bit about this question on HackerNews: I had some thoughts about some of the points that were mentioned in the conversation.

In a large corporate environment, the decisions that are made at the highest levels can be bewildering, seemingly inconsistent and out of touch with operations on the ground.

In the realm of engineering team organization, a one-size-fits-all approach simply doesn't suffice.

People are complicated.

I spend a good bit of time in conflict resolution - either my own, people on my team or between groups.

Projects pulsate with life.

The topic of predictability in software engineering is one that comes up often at work.

In the past several months I’ve seen the affect that poor communication has on the efficiency of an organization.

With my website (which uses Notion as its CMS), I ran into a problem caused by a quirk of Notion’s API.

I’ve been tinkering with Notion as a CMS for the last few days and decided it checked all the boxes for a simple site like my own.

Let’s start by defining the difference between A/B testing and feature flagging.

With communication as difficult as it is, keeping your development team as small as possible yields much better outcomes for a lot of reasons.

Ansie is a rich-text terminal rendering package I started this package out of frustration of not having a relatively simple way to output rich text without relying on direct terminal codes which is difficult to parse with the human eye.

While there are some examples of how to setup a slackbot that runs on Lambdas, there wasn’t much available for running in an edge environment.

I recently decided to switch an internal tool from NextJS pages router to app router.

I recently made the switch from Notion to Markdown for a couple of reasons.

I am most happy when I play even a small part in someone’s success.

Organizations leak knowledge - it’s the nature of organizations.

The decision to build a solution vs. buy one is a point of some contention.

I’ve started to put more thought into the notion that’s been brewing in my head for a while.

The in-between hosts necessity but not much in the way of excitement.

My son was saying today that this project he’s working on I realized that this is actually a good thing.

Your co-workers are your partners.

I have been managing a mix of onsite and remote teams now for about seven years and prior to that worked with remote teams as an engineer for about twelve years.