About
Very Smart Parrot
Practical AI engineering — from assessment to production.
We build AI agents inside global enterprises — not demos, not proofs of concept, but systems that handle real decisions with real consequences. The distance between a working prototype and a production agent is where most projects die. It’s also where we do our best work.
This site exists because we got tired of the gap. Tired of AI content that stops at the impressive demo and never mentions what breaks the moment you scale it. Tired of hard-won lessons staying trapped inside corporate walls.
So we write about what we actually see: which architectures survive contact with messy data, how evaluation works when there’s no clean benchmark, what it takes to trust an agent enough to let it run unsupervised. The problems nobody talks about because they’re not glamorous — but they’re the ones that matter.
Everything here is rooted in daily production work with LLMs and AI agents. The guides, the research, the open-source tools — all of it earned, not theorized.
Updates
Subscribe to get an update about new content by email.
Contact
We read every message. Whether it's a question about something we published, a correction, or just something you wanted to share — we'd like to hear from you.