About Applied GenAI

Practical, opinionated recipes for building real things with generative AI — less hype, more patterns that survive contact with production.

There's no shortage of generative-AI content that's either breathless ("everything has changed!") or abstract ("attention is all you need"). This site aims at the gap in the middle: the concrete, hard-won patterns you actually need when you sit down to build — how to make an agent reliable, how to keep a model honest about intent, how to give it memory that doesn't rot, how to manage prompts like the assets they are.

What a recipe is

Every recipe follows the same fixed structure, on purpose. You'll always find the problem it solves, the forces in tension, an architecture diagram, the steps, real code, the trade-offs, the pitfalls, and a checklist to adopt it. The consistency is the point: once you've read one, you know exactly where to look in all the others.

And every recipe is meant to be runnable. Where there's code, it lives in the companion cookbook — a real, tested Python package you can clone and learn from: applied-genai-cookbook.

Who writes this

Written by Mohit Mittal. (Edit src/pages/about.astro to tell your story — what you build, what you've learned, and how to reach you. A short, specific bio does more for credibility than a long one.)

Follow along

New recipes show up in the RSS feed. The code is on GitHub. That's the whole operation — no newsletter funnel, no popups, just recipes.