Hey, I’m Khaled

Software engineer building in public. Writing about code, side projects, and whatever else is on my mind. Currently heads down on TaleTribe.

Going Invite-Only: What Slow Growth Taught Me

Everyone tells you the same thing about a new product: remove friction from signup, don’t add it. Open the doors, let anyone in, worry about scale later. I went the other way on purpose. I want TheTaleTribe to feel a little exclusive, so instead of a signup form you get an invite system — you actually apply to become an author. And at least for now, I’m not approving everyone who applies. Before I open the gates to the world, I want to seed the platform with a few real stories and partner with some actual authors, so that whoever shows up next walks into something that already feels alive instead of empty. That’s the romantic reason. There’s also a colder, more practical one. ...

May 15, 2026 · 4 min · 698 words · Khaled

The Day Chat Got Faster (and Cheaper)

For a long time, chat on TheTaleTribe worked in a way that embarrasses me a little now. Every single time you sent a message, the backend would go read every chapter, every character, every place, every plot note attached to your story, and ship the whole thing to the model as context. It worked. It felt like it was reading your whole book because, technically, it was. And for the handful of short test stories I was using during development, it was fine, fast enough, cheap enough, nothing to worry about. ...

April 30, 2026 · 4 min · 805 words · Khaled

Teaching the AI to Remember Your Story

What I really wanted was for the AI to actually know what’s going on in your story. Mention a character and it should already know who they are and what they want — assuming you’ve told it, and assuming it doesn’t just make something up on the spot. That sounds simple, but getting it to hold a whole novel in its head without either forgetting things or hallucinating them took me longer than almost anything else I’ve built on TheTaleTribe. It’s easily the most complicated part of the whole project, and it’s the piece I’m probably proudest of. ...

April 5, 2026 · 5 min · 900 words · Khaled

Keeping the Lights On: How I Keep AI and Infrastructure Costs Sane

This is a passion project, and it’s just me paying for it. So the finances are something I have to take seriously, even now, before there’s really any traffic to speak of. The nightmare isn’t failure — it’s a weird kind of success: a story gets shared somewhere, real users actually show up, and I wake up to a bill big enough to make me regret ever letting people in for free. Long before that could happen, I started building barriers. ...

March 10, 2026 · 7 min · 1281 words · Khaled

Bring Your Own Key: AI on Your Terms

The Ask I Didn’t Want to Make I realized pretty fast that the only way to keep the real power users happy was to let them bring their own key. It’s not the most elegant answer, and I know it. Asking a non-technical writer to go generate an API key on some other website and paste it into mine is a genuinely big ask. But I don’t have investors, and if I want to keep my costs survivable, some things have to give. So for the people who live in this app, I built the BYOK flow. ...

February 20, 2026 · 3 min · 506 words · Khaled

Why I Bet on Firebase and GCP as a Solo Builder

From Side Project to Full-Time Obsession I started building TheTaleTribe a long time ago. Initially, it was just a side project—a playground for experimenting with new AI models. While there are plenty of public benchmarks out there, I prefer having my own. Building actual features became my way of stress-testing the latest and greatest LLMs to judge their true capabilities. The project started with Gemini. If you look at the first video on my YouTube channel, you can see how rough the app looked at the start, but the core idea was there. This was right around the time ChatGPT was first becoming popular. ...

January 25, 2026 · 4 min · 735 words · Khaled

TheTaleTribe: Origin story

The TV Kid Era I wasn’t a big reader growing up but I did watch a lot of TV, spending most of the 90s watching whatever was on—and I mean whatever. This was before Netflix, before YouTube, before any of that. We just watched what was on TV, and honestly, I think that made me consume more of a variety than I do now. There was no algorithm, no endless scrolling—just whatever happened to be playing. Then social media came along and made it even worse. Having too many options and too much noise everywhere you look just makes it harder to sit down and actually enjoy something properly. ...

January 4, 2026 · 4 min · 790 words · Khaled