SKHELL Tia Zanella | 2025 Retrospective

For me, 2025 was a year of intense personal challenges. It forced me to step away from the noise of society, hit pause, and engage in some deep reflection. It was a time to rethink what really matters in life, in work, and in the projects we spend our energy on.

Throughout it all, I kept building, refining, and learning the same fundamental lesson in different ways: systems don’t care about intentions. They care about interfaces, constraints, feedback loops, and how they behave under pressure.

On AI and hype cycles

AI is everywhere now. I don’t hate it; in fact, I see its potential and use it as a powerful tool in my daily workflow. What concerns me is its thoughtless deployment, and the all-too-familiar corporate playbook playing out once again: speculation over substance.

A few years ago it was the cloud, then cybersecurity, now AI. The playbook never changes. The same “smoke-sellers” riding the hype wave, the same pay-to-play “magic quadrants” rating software, the same suits selling slide decks without ever having run a terminal command, and the same enterprise vendors peddling overpriced garbage that increases complexity instead of solving actual operational problems.

What’s worse is the persistent feudal mindset. Too many engineers and managers are still protecting their tiny paper kingdoms, hoarding knowledge, and gatekeeping behind expensive paywalls and certifications. Instead of learning, failing, repeating, sharing what works, and building on each other’s progress, the industry focuses on extraction: monetization first, proprietary lock-in second, and actual, diluted innovation a distant third.

It’s exhausting watching this loop repeat while ignoring the incredible open-source tools we already have. Imagine what we could deliver if we stopped treating technology as a zero-sum land grab.

What I built

In 2025, I shipped Net Commander, a network operations toolkit for VS Code focused on removing friction from repeatable engineering tasks.

I broke things. I simplified systems that were getting way too clever. And I kept returning to one core principle:

Systems. Tradeoffs. Reality.

I built small utilities and optimized workflows because I am convinced of this: the best automation removes human pain without introducing new, unpredictable failure modes.

What I learned

Speed without guardrails simply scales the blast radius.

The faster your deployment pipelines move, the more you need structure not annoying corporate bureaucracy or AI agents that pretend to make coffee, but deterministic constraints that prevent runaway system failures. This showed up repeatedly in 2025: in configuration management, in continuous integration, and in every system that prioritized velocity over resilience. And with the current rush to inject AI into everything? The chaos is only accelerating. We need structure now more than ever.

Looking ahead to 2026

After years of working inside proprietary corporate walls, I’ve decided it’s time to turn my personal notes and long-held ideas into public tools. I am investing significant time building something meaningful for the open-source community that I will unveil in 2026, alongside a few other side projects I am genuinely excited about.

Thank you

If you read a post, opened an issue, sent a message, or challenged an idea of mine this yearthank you. It matters more than you think.

Happy New Year,
Tia