Posts

In our platform, we run a multi-step AI agent that generates content for customers in the sports world. The agent is never fixed: we change the underlying models, reword prompts, restructure workflow steps, adjust the agent logic. Any of those can quietly make the output worse, and for a while the only safeguard was a human, reading generated content and forming an opinion. That doesn’t scale at all, highly subjective and misses slow drift completely. ...

After trying many personal productivity apps and even attempting to vibe-code my own ideal one, I switched back to good old Todoist. However one recent discovery make it significantly more useful to me: official todoist-cli. ...

I built an AI code reviewer that actually fits into our workflow. It’s a GitLab CI pipeline that reviews merge requests automatically using Claude Code and posts findings as review comments. ...

I’m worried that adding AI to our code review pipelines might actually make our code quality worse - not because of the AI, but because of us. ...

It’s been quite a while since my last post here. Honestly, 2025 was an incredibly overwhelming year for me professionally. I finally found the time to sit down and write this slightly delayed wrap-up, reflecting on what happened and where it’s all heading. The biggest shift for me was not just that AI tools became better, but that my role started changing around them: from writing more code myself to organizing work, context, quality, and small agent-driven workflows. This post is my attempt to make sense of that shift: the productivity gains, the stress, the documentation paradox, and why I think 2026 will be less about hype and more about control. ...