Indie Diary #7 — Recharging batteries

It’s been a while since my latest diary entry. This entry will be both about Omee, but also about why I wasn’t around the past month.

Omee’s beta learnings

The Omee beta launched on February 29th. Since then, the usage of the app has been pretty low.

And I can see multiple reasons why:

• HomeKit is a very small niche.
• White page syndrome: people do not know what to ask the app.
• Hard to trigger: even a widget tap is less efficient than a “Hey Siri” wake word.

Aside from this lukewarm reception from the public, Omee suffers from bigger issues that need a lot of work to solve:

• The UI was lacking ways to customize the generated scenes. This would require an almost entire rewrite of the current backbone of the HomeKit layer of the app, a colossal task.
• The communication with HomeKit was sometimes flaky. I hoped to solve this with the above work.
• The LLM would sometimes hallucinate. Improving the LLM reliability became harder and harder, with mixed results.

You might think that I set my quality expectations a bit too high, but I value the quality of the products I ship a lot. And I couldn’t assume to make people pay for something that wouldn’t work as I’d like it to work. If most of Omee would be “okay”, it was, for me, far from a release as a premium product.

Taking a step back

In parallel, while working on Omee and acknowledging all the personal failures mentioned above, I was under a lot of pressure from my employer since December. Firstly, my employer pays my bills. Additionally, I have the chance to do high-value and very interesting work for them.

That’s why I started shifting from working on Omee to working a bit longer for my employer. The idea was to remove weight from my shoulders, alleviating some of the current pressure at my daily job, and focusing on something that actually works.

I was feeling guilty about not working on Omee anymore, but my motivation for this project felt very low, and I couldn’t muster the very little available energy I had for it.

It was a better decision to pause the project entirely.

It’s still frozen as of today. It might resume depending on what’s announced at WWDC 24. It could be a very good use case for an on-device LLM. We’ll see.

Preserving my own health

After the project froze, I had no energy to start a new project, nor continue my presence on social networks. That’s mainly why since the beginning of April, I’m pretty much nowhere to be found.

Preserving myself is one of my biggest priorities. The missteps with Omee, combined with my heavy daily workload, the content production for social networks and the blog, and a few events in my personal life, became too much to handle on a daily basis. Scaling back almost all of my indie tasks workload for a month was my solution to preserve myself.

Good news, it worked! This blog diary is my comeback.

Taking time for myself after work was a step to rebuild confidence and prepare myself for another project before WWDC. I have a few candidates. I should be able to start working on them in the next few days.

I see a lot of similarities with the Omee situation when watching Rabbit r1 or Humane AI Pin reviews. The main difference is that I’m not building a physical product, and I have no investors to satisfy. Therefore, I have the luxury to choose not to ship.

The comeback

I’m going to ramp up my presence on social networks and open-source projects from now on. I have a lot of things to do, and I don’t want to burn myself out by doing too many things too quickly.

Therefore, now is a good time to build a small project. I want something easier to solve and quicker to ship. I’ve yet to finalize the choice for that project, but you shall hear from this soon enough.

Also, I’d like to release my library ApiKit. It’s a small abstraction around HTTP calls that leverages Swift features and type-safety.

Finally, I’ve been invited to talk about API security on mobile apps at a French event: Mobilis in Mobile. This talk will be in French, but it’s a good opportunity for me to step back into the indie community.

Let’s do this!