.png&w=3840&q=75)
The Claude subscription mess, explained
Anthropic just split Claude subscriptions in two, and if you build on top of the SDK, your costs are about to change.
Adrian OleszczakAll Articles
Blog
Your friendly guide to AI and tech.
10 articles
Adrian Oleszczak
Volume is preventing your company from scaling and AI workflows are the only way out
Scaling fails because of volume, not complexity. Discover how AI workflows can handle repetitive operational noise, eliminate the "human shield" layer, and protect your roadmap.
Pawel Bazyluk
Why Architectural Kata Matter?
Most architects rarely get the chance to truly practice architecture. They often work alone, receive feedback months later, and struggle to clearly explain their ideas to non-technical stakeholders. Architectural Kata create a safe environment to train the core muscles of architecture: problem framing, trade-off decisions, system thinking, communication, and risk anticipation.
Piotr Filipowicz
Team Topologies for AI Agents
Team Topologies was designed to reduce cognitive load for human teams. But what happens when some of your team members are AI agents? If we don’t define clear boundaries, structure, and ownership, agents won’t fix the chaos. They will simply scale it.
Piotr Filipowicz
The end of the manual grind: AI workflows are the ultimate human upgrade
You hired your team for their strategy, not to act like tired robots. Learn about AI workflows and how they take over repeatable processes, eliminate costly errors, and free your best minds to do the work that actually grows your business.
Pawel Bazyluk
ADRs in the Age of AI
In an AI-augmented world, architecture decisions can no longer be just documentation for humans. When coding agents operate inside a single repository while architecture spans an entire ecosystem, ADRs must become layered, structured, and machine-consumable. This article outlines a practical model to keep local teams, cross-repo systems, and enterprise constraints aligned for both humans and AI.
Piotr Filipowicz
Pottery Class Paradox
The article uses the Pottery Class Paradox to show that quality often emerges from repetition, feedback, and iteration rather than from chasing perfection upfront. It connects this idea to evolutionary architecture, where systems are designed to support continuous, guided change protected by fitness functions.
Piotr Filipowicz
AI writes code better than me
A big shift is happening in software development. AI already writes code better and faster than many developers, but it still does not understand the bigger picture. It struggles with architecture, trade-offs, and long-term decisions.
Piotr Filipowicz
Extreme Go Horse Manifesto Meets AI
For those who don’t know it yet, eXtreme Go Horse (XGH) is a satirical “methodology” created as a parody of bad software development practices. It glorifies speed over thinking, hero coding over architecture, and blind optimism over responsibility. Funny? Absolutely. Dangerous? Also absolutely, especially when combined with AI. Today we are witnessing the birth of a new anti-pattern: AI-powered XGH.
Piotr Filipowicz
How I One-Shotted This Blog Using the Ralph Wiggum Technique
I one-shotted prompt-pals.com in about 90 minutes by stepping away from manual execution and letting a "Ralph Wiggum" loop take over. It’s a shift from just chatting with AI to orchestrating a persistent agent that grinds on a project spec until the job is actually done. Here is the breakdown of the autonomous workflow and the exact XML spec I used to make it happen.
Adrian Oleszczak