About

I’m a working software engineer writing from inside real projects and this is my learning log.

I'm learning how to use agentic AI as a coding companion — not as a replacement, not as an oracle, and not as a productivity miracle. Just as another thing in the loop while I think, design, debug, and occasionally get confused. 

The metaphor here is the rubber duck: a thing you explain your thinking to until something clicks. Sometimes the duck helps. Sometimes it confidently leads you astray. Either way, the act of explaining turns out to matter more than the answers.

This is not a tutorial site.
This is not a product blog.
This is not a promise that anything here is correct.

What you’ll find instead:

• short notes from real workexperiments that half-work
• failures that taught me something
• evolving opinions about agents, tools, and boundaries

Posts are written as they happen, not retroactively cleaned up. Ideas may change. Old posts may age badly. That’s intentional.  I may even change the posting rules, but for now they're right here.

If any of this is useful to you, great.
If not, that’s fine too — the duck and I will keep talking.