About Me

I’m Jishan Baig. I write here about building software around AI as a runtime primitive, not a bolt-on.

Most production AI work isn’t about the model. It’s about the boring stuff that surrounds it — schemas, retrieval contracts, eval harnesses, observability. The pipeline you can trust at 3 a.m. when production has drifted. The system that fails loudly instead of quietly. The architecture that lets you swap the model next quarter without rewriting the application.

That’s what this blog is about.

You’ll find three kinds of post here:

• AI-Native Architecture — design patterns for systems where the LLM is load-bearing, not a feature
• Agentic AI — what changes when your software starts making decisions
• Architecture Patterns — SOLID, clean code, and DDD applied across classical, cloud, and AI systems

The model is the easy part.

I’m on LinkedIn if you want to push back on a post or talk shop.