Model capability is only half of what you get from a coding agent. The other half is the harness: the loop, the system prompt, the tools, the permission rules, how it remembers things, how it shrinks
A engineer sharing hands-on experiments with LLMs, agents, and inference—what worked, what failed, and the trade-offs.
I’m a engineer working across generative AI, LLMs, and data engineering. I enjoy building practical systems and understanding what happens beyond the demo—how they perform, where they fail, and which trade-offs matter. My focus is straightforward: explain what I tried, how I tested it, what worked, and what I’d do differently. Expect code, honest observations, and practical takeaways—not just feature lists or benchmark headlines.
Markdown supported · ⌘↵ to post
No responses yet.