Day 79 of building Gradiora (ai language tutor): Some coding back on the menu! With gpt-realtime released I want to give some users the chance to a/b test vs 4o-realtime. Almost there after an hour or so today but a bit confused by changes to transcripts in session creation.

Man, Whop/App Mafia really knows the importance of personalization in marketing

Day 80 of building Gradiora (ai language tutor): First beta testing active gpt-realtime instead of 4o-mini-realtime/4o-realtime. Definitely an upgrade and looking forward to deploying to all users once gpt-mini-realtime is available. Also considering adding a higher priced tier

Tiktok cheat code found: talk to kids. My 12yo son showed me what he thought works and he quickly got a 200k+ views faceless post. Good upgrade from a few hundred...

Day 81 of building Gradiora (ai language tutor): Back to marketing as main focus. Today was learning day to see what works. Obvious things like right hook and right music. Easy to say but hard to do though. Getting help to do this through September.

Day 82 of building Gradiora (ai language tutor): Consolidated user feedback today to set the plan for next 2-3 updates. Have tested a few tools but end up reverting back to a simple list of bullets in a doc. Feels like everything else slows down the process / becomes restricting

Day 83 of building Gradiora (ai language tutor): And so it happened, had a day where I didn’t have time at all to build or market. Frustrating but simply too much else today. Back at it tomorrow.

Day 84 of building Gradiora (ai language tutor): 30 min coding and 30 min marketing today. Made good progress on 2.0 branch and one new user secured. Hoping to set aside a good slot this weekend for a longer session and push next update by Monday.

Day 85 of building Gradiora (ai language tutor): Mainly coded today. Should focus on marketing but have a clear view of what I want released first so pushing ahead for a few more days.

Day 86 of building Gradiora: Finally moving on to marketing. Starting with guidance from Ernesto Lopez now have 2 agreements in place to generate and post content. $1/k views seems w. $1k cap seems like acceptable model. Initially smaller scale to test it out.

Day 87 of building Gradiora: Fun milestone today when I noticed first actual payout to bank account. No longer just a number on a dashboard! Will move from numbered daily updates to just posting as I go. Daily summaries not the best format. Was good to get started though.

The lack of an iPhone 17 Plus is so frustrating. See no need to upgrade to the pro max. Don't see the any real value with the Air for. Don't think I can leave my iPhone 15 Plus for a smaller screen.

Marketing efforts continue and some payoff in the form of MRR growth gives new energy. Variable costs and cursor subscription now covered.

Decided to take a couple of days off from the evening coding and marketing sessions. Still gained one new paying user. Gained both energy and clarity from a few days away from it. Now back at it with the iOS 26 release next.

Spent the evening building. No, not like that. The time honored tradition of putting together an Ikea bookshelf for the home office. Best recently read book put in it? Amor Towles' A Gentleman in Moscow. Next step, move more books to the office...

Goals for the coming two weeks: - Get iOS26 ready version of the app out - Break through $100 mrr (at $65 so still 50% growth)

I get doing the "podcast tour" when promoting a new book but when my favorite Monday pods go this similar it is a bit jarring.

Tiny milestone but passed $100 now with the first app

Ok, this was actually a really great idea. Really appreciated being able to dive in and ask a couple of questions that lingered having listened to the podcast and read Karpathy’s post.

gpt-realtime-mini, finally! Time to get going. That quality at that price point truly makes several ideas viable. BRB