I didn’t write for nine months. I was shipping.

Last October I started a weekly newsletter here. I lasted three weeks. Then we started a new application at work, and this place went quiet.

The new thing was a mobile app. I had not shipped one before. I had spent years on backends, web, a desktop tool. Not App Store, not Play Console, not a phone in someone’s hand on the tube.

The stack is React Native on the client, .NET on the server, Kubernetes in production. Voice in a live session is a different problem than a form that saves to a database. You care about the call connecting, the delay, what happens when someone joins late, what happens when the network dies in a tunnel. I had to learn that.

I also had to learn the stores. Listings, screenshots, review notes, privacy questionnaires, version after version. Purchases: Apple, Google, a trial that starts, a receipt that fails, a subscription that renews when you didn’t expect it. I configured all of that myself. Same for the rest of the release path. There was no mobile team to hand it to.

The product is Daily Talking. People who want to practice speaking English join a small live voice group, four to six, no teacher. They can listen first. If there is no live session, they can use AI Talk, a few minutes of 1:1 speaking, then a short recap. Subtitles and translation sit under the call. Sometimes the group plays trivia so the session doesn’t die in silence. Two moderators work on getting people who don’t speak English yet to actually talk, and on keeping groups alive. I manage that too. I look at PostHog, see where people stop, decide what we build next, and I do the marketing: the public blog, ads, people who might show the app. One product, one engineer. It is tiring. I am not complaining. The more I understand how people actually use the app, the more I want to keep going. I know I neglected this newsletter while that was happening. Things are starting to settle now. I will write here often, even if it is not every week.

If you want to try it: https://dailytalking.com

iOS: https://apps.apple.com/app/id6757860877

Android: https://play.google.com/store/apps/details?id=com.volosoft.dailytalking

If you have shipped a mobile app, I want to hear from you. I am looking for feedback, in private is fine. I am still early at this part of the job and I would rather learn from someone who has already shipped than keep guessing.

Building this is not the same job as building a framework for developers.

With ABP, your user is a developer. They can describe what broke, they open an issue, they often keep using it while they wait. That does not make the bug cheap. It makes a conversation possible. With an app like this, if the first session is bad they delete it before you wake up. You often don’t get a report. You get a flat line in PostHog, or a one-star store review that says “doesn’t work.”

I learned a lot from both. The years on the framework were not wasted on this. Permissions, background jobs, versioning, thinking about how someone else will use what you shipped: that came with me. Mistakes still matter there. A developer files an issue and usually stays. A person who opened a speaking app once may never come back. What I had to drop was the habit of saying the module handles it. Here if the subtitle lag makes the session unusable, I change it. If AI Talk sounds clever in a demo and stupid after three minutes, I change it. If people install and never speak, that is on the path I designed.

I still work on ABP. I still run the occasional training. Studio’s AI coding agent is another story. I’ll write that later.

Anyway. This was just an update on what I’ve been doing. I have more to write, and I will. We’ll talk again soon, more often than this year.

See you in the next one. 👋