In 2026 one of the skills that every software engineer, developer or architect will need is the knowledge of quickly identifying the root cause in (unknown) code. This will be a differentiation factor, as more and more code will be AI generated and developers will not know all portions of the code. Being able to visualize the data flow and/or the code flow in your head will proof you to be human and show that your brain is still more capable than any AI coding assitent. Experience and mistakes matter - they set you apart from what's happening in our industry. #SoftwareDevelopment #AgenticAI #BuildInPublic

Is "platform as a product" a mindset, or a slide we show once and never fund?

We test our code relentlessly. When did we last test the pipeline that ships it?

The agent is always confident, even when it's wrong. Is the new skill knowing when to doubt it?

Did multi-account, multi-region really buy us resilience, or just a bigger surface to misconfigure?

Does your platform make the right thing the easy thing, or just the only thing?

📊 App Stats Update (last 30 days) 🚀 393 new installs across 6 apps 🔄 300 app updates 📱 1,530 active devices (Android) 🍎 ~25 daily active (iOS) 📈 +5.6% installs vs last period • Gem,Ore & Struct Finder for MC: 1,487 installed, ~25/day iOS, +385 new, 223 updates, 3.3★ • HC VfL: 15 installed, +6 new, 68 updates, 5★ • Nexus Share - Social Scheduler: 14 installed • Pegasus Galaxy: 6 installed, +2 new, 7 updates • Train with Joe: 4 installed, 2 updates • Unscratch the World: 4 installed #indiedev #appdev #buildinpublic 🍎 lckhd.eu/csKqc2 🤖 lckhd.eu/MbGsPt

Every team has one pipeline step nobody dares touch. What's yours, and when did it become untouchable?

I'm speaking at the #AWSUserGroup Mannheim on Wednesday, would love to see you! RSVP herE: lckhd.eu/09KOUN

Oh yea, doing that #awscommunity

After following Tim follow me. I have a similar profile

Is "harness engineering" a real discipline forming, or just all of us improvising in parallel?

We tag, dashboard, and alert on everything, and still find out about the bill on the 1st. Why?

We build platforms so teams stop repeating work. Then every team customizes until it's bespoke again. Why?

Is a fast pipeline a real win, or a vanity metric we optimize because we can?

Sounds like a really strange move

We used to know why the code worked. Is understanding becoming optional, and are we okay with that?

Is our cloud automation preparing us for failure, or only for the failures we already understand?

Where's the line between enabling teams and quietly centralizing the decisions they used to make?