Summer 2025 - July Update

It's been a while since I've written here, so I figured I would do a post to update you on everything I’ve been up to since last time you heard about me.

In this issue:

• Moving to a different country and continent

• AI-assisted programming with Claude Code

• Quick calorie and fasting tracker prototyping

• Productizing my consulting services

• Expo and React Native thoughts

• Reading two new books

• Update on my summer roadmap

I've always dreamed of being an indie developer. The reason I finally pulled the trigger is that I decided to move to another country, and even another continent.

This allowed me to peacefully leave my day job and really try indie hacking.

I moved from near Lyon and Geneva in France, to Tangier, Morocco.

Indie hackers usually move to Asia or Southern Europe, sunny places with low cost of living and friendly to entrepreneurs.

I am happy to tell you that Morocco (especially the North) is also on the list, and I frankly don't understand why everyone is sleeping on it:

• The population usually speaks French, English, or both

• It's sunny, hot (but AC is widespread), with tons of beaches, both on the Mediterranean and Atlantic

• Cost of living is low

• It's only a couple of hours away from Europe by airplane or a quick 1h ferry to Spain

• You can quickly drive to Ceuta which is literally a Spanish city on the Moroccan territory (enclave/exclave)

• There is quick and cheap fiber and cellular connectivity everywhere

• There is a vibrant tech and intellectual community

Since I don't see anybody documenting life in Northern Africa, I'll try to raise awareness on social media and show what life here is like, should people be interested.

If you’ve read my build in public series on my app Summiqo, you might remember that I extensively used Cursor to build it.

It was great, but the AI world is moving really fast. Since then, it feels like dozens of trends have come and gone. In the meantime, I tried to filter through the noise and I've stuck (at least for now) with Claude Code and a Max subscription (to get access to Opus 4).

Using Claude Code is really one of the best ways to have AI help you write code. The fact that it's a simple terminal app might sound unappealing, but it is exactly what makes it so powerful. It removes all the cruft, bells, and whistles, and it's essentially going back to the early days of computing.

It allows interacting with this new paradigm of programming (natural language with LLMs) through the simplest interface ever imaginable.

Every time I want to make a substantial edit in my projects, or do a quick prototype, the very first step is always Claude Code.

While it's true that the learning curve is a bit steeper compared to other GUI tools, in the end it's really rewarding, giving you much more power and flexibility.

Anthropic published a nice course to learn Claude Code on their online academy. I suggest you follow it if you don't know where to begin, it's a great way to get started.

I’m trying to be careful about how much I eat and I do intermittent fasting for health benefits.

Historically, calorie tracking has always been relatively painful, since what we eat comes from so many different sources. I downloaded Cal AI to see if it would fit my needs. And I immediately encountered two issues for the usage I envisioned:

• The calorie tracking aspect has a lot of features, and that makes it complicated for the basic use that I envisioned

• There is no way to track fasting duration between meals

So I ended up making a quick prototype of my own.

This is one of the novel aspects of AI-assisted programming (in this specific case, you could call it vibe-coding).

I almost one-shotted the prototype with Claude Code, in order to get something and be able to use it myself. From there, I can refine my vision and then ultimately decide on what I want to build and if I want to build it at all.

Since then, I've been using both Cal AI and my prototype to test how well I can estimate calorie intake based on pictures and descriptions. The results are promising, especially considering that I haven't optimized the process at all yet.

It's been a few weeks now, and the jury is still out. I think I'll decide during the last part of the summer since I already have some projects I need to finish before starting a new one.

While I’m growing as an indie developer, it doesn't pay the billets yet, and it might not for a while.

So I rely on some consulting and freelancing as a source of income.

In order to streamline my freelancing work, I got inspired by the DesignJoy approach.

It initially started as a simple and cheap subscription service for unlimited design work, and now has grown into a bigger business.

So I made the following website: swiftswiftapps.com

The gist is as follows: I offer my freelancing services, especially building MVPs for a very competitive price, extremely short shipping delay, and a small monthly maintenance fee.

This way, I turn my services into a product, I industrialize my building process (assisted by AI of course), and I leverage my experience in Swift / iOS / macOS / mobile. This is my attempt to adapt to the freelance market in this day and age.

I think what I offer my clients is a very good deal.

They get a fully functional MVP, plus market research and ASO, shipped to the store for a few thousand dollars, and unlimited maintenance for a few hundred dollars a month.

If I wasn't a developer myself and I wanted to make an app, this is definitely an offer I would consider.

Speaking of my consulting business, I’m focused on Swift and Apple technologies, but when it comes to cross-platform needs, my experience is in React Native.

That's why I wanted to play with it a little bit, to see how far it has come in 2025.

First, the good: Expo. This set of tools provides a really great developer experience. I really wish we had something that efficient in Xcode, where SwiftUI previews have been painful ever since they appeared.

TypeScript is also quite nice to work with, and is on par with Swift on most use cases.

However, I couldn't shake the feeling that even a simple app felt out of place both on iOS and Android. Achieving a good design is significantly harder, especially if you want it to adhere to platform-specific UI paradigms.

This is an option I will keep in my arsenal, but only for what I consider it good at: simple MVPs that don't require an extraordinary design and are constrained by time and budget.

I've acquired two new books that are probably going to be bangers in the indie hacking space.

The first is My Indie Book by Tony Dinh. Tony is really well known in the indie hacking space, and he's consistently provided honest and useful takes, so I’m excited about this book.

The second is Everything but the Code by Paul Hudson. If you’ve ever written more that a line of Swift, there's no way you haven't benefited from Paul's work. He's the guy behind hackingwithswift.com and so many books.

The approach Paul took is a bit different, he invited an expert for each chapter of the book, and I can tell you that indeed those are all successful indie app makers.

Regarding my summer roadmap, I didn't make much progress because of moving, but here's the current status:

I plan to bring significant updates to my transcription and summarization app.

First, I would like to use the new Apple local APIs (SpeechAnalyzer and FoundationModels), and only keep a fallback on Azure and AWS Bedrock.

This will allow me to make the app free with a tip jar, or very very cheap, in order to gain more visibility, popularity, and increase my reputation.

Also, my app is still missing an iOS UI, which I planned to include from the beginning.

I will continue writing here on various topics, probably as I encounter them along the way.

I would like to participate in RevenueCat's Shipaton. Ideally I'd like to enter with the iOS version of Summiqo, but I don't know if it really qualifies as a new version according to RevenueCat's terms.

If it doesn't qualify, I still have my calorie and fasting tracker app, or even some other ideas that are on the back burner:

• An app around learning to read Arabic

• An app for Tangier visitors

• An app for people traveling by car from Western Europe to Morocco. Unfortunately, crossing Spain comes with a lot of potential issues: theft, empty areas, traffic jams, etc. The idea is to have a crowdsourced map to stay safe and have a companion helping this road trip.

What are you working on? Do you have feedback on any of this? I’m eager to hear your thoughts.