2025 hit different.

2.6x more shipped than 2024. Measured by commits, projects launched, and hours logged. Checked my “2025 predictions” from last December - 80% hit, some overshot. Writing down what you want actually works.

What happened

Started the year with the usual exams. Boring.

January - palora. First SwiftUI app. Realtime color extraction from camera feed. Nothing crazy, but I wanted to learn the platform and I did.

February - Published our second book: Probleme si Teste Grila de Informatica for Romanian high schoolers prepping for UBB admission. Physical product, actual distribution, people using it to get into university.

February - Optima. AI that roasts your workouts. My first real 0→1 product. Made some tiktoks with zero experience, got ~1k views total, converted 7 actual users.

Then it died.

Here’s what I learned: the code took a weekend. The App Store Connect bullshit, the icons, the screenshots, the metadata - that took longer. And marketing? Marketing is a full-time job pretending to be a side task. Those 7 users taught me more than any tutorial. I talked to them. I watched how they used it. That’s the game.

May - Max and Aavash from Deta came to Romania for a few days. Always good to hang with the team IRL.

June - Joined clujhouse after meeting Andrew earlier in the year. Started at Anara shortly after. Six months in, feels like two years. We ship fast, I actually use the product daily, and I’ve been polishing UI and pushing features since day one. 2026 is going to be bigger.

August - Traveled to the US for the first time - San Francisco. Reading my 2024 review I realized I actually wrote this down as a goal and forgot about it. Somehow predicted it. SF is a different world. If you want to build something, that’s the place. Even just being there and absorbing the culture gets you further than you’d expect.

December - Made hive, a small open source CLI for running N agents on N tasks in parallel without them shooting each other in the foot. Still a work in progress but the core idea works.

December started vinculum

I’ve tried to build this three times before and failed. The idea: education doesn’t scale because good teaching is 1:1 and we’ve been stuck with 1:30 classrooms for centuries. LLMs change that equation. One tutor per student, infinitely patient, available always.

Super early. If you want to try it, reach out. Launch planned for 2026.

The numbers

Activity	Time
Mac Screen Time	8:38hr on average [3,151 hours in total]
iPhone Screen Time	0:52hr on average [316 hours in total]
University	2hr on average [280 hours in total]
Driving	240 hours [12,431km]
Gaming	200 hours [Battlefield 6 or Arma Reforger]
Gym	1hr on average [163 hours in total]

I track most of this passively. Screen Time, Apple Health, apps with year-end wrapped features. If you’re not measuring you’re guessing.

What Actually Stuck

On building:

• Ship first, optimize never (or later). The best architecture with zero users is worthless.
• Do the simplest thing that could possibly work. You can always add complexity. You can’t always subtract it.
• Research is procrastination wearing a lab coat. Build something, watch it break, fix it.

On marketing:

• It’s not a side quest. It’s the main quest. Code is the easy part.
• Nobody is posting. Everyone is consuming. The bar for standing out is low if you just show up consistently.

On learning:

• The smartest people are self-taught even if they went to school. Credentials get you in the door. Skill keeps you in the room.
• You can learn anything. The “too late” thing is a cope.

On mindset:

• Your plan is always smaller than what’s possible. Most limitations are self-imposed.
• You get what you believe you deserve. Sounds like woo but watch how you negotiate, how you price yourself, how you ask for things. It’s there.
• The cold water doesn’t get warmer if you jump late.
• Everyone is figuring it out as they go. The people who look like they know? Better actors.

Thanks for sticking with me this year. You the real one.

2026

This year is going to skyrocket.

More vinculum. More shipping. Less overthinking.

10x the output. 1000x the results.

See you next year.