AnupTechTips began with a small frustration. So much writing on AI and large-scale systems either floats on the surface or buries the real ideas under jargon. I wanted the opposite: clear, plain-language writing on AI architecture, large language models, data science, and the craft of building systems that hold up under real load.
What you’ll find here
Everything here is hands-on and meant to be used, not just read:
- Step-by-step tutorials on LLMs, AI agents, and modern AI tooling
- Deep dives into data science and exploratory data analysis
- Patterns and trade-offs for building scalable, reliable systems
- Practical Python tips and the workflows that make engineering calmer
About the author
Hi, I’m Anup Das. By day I work as an Applied AI Systems Architect, which is a long way of saying I design the moving parts that keep AI and backend systems running reliably at scale: system design, GenAI infrastructure, and a lot of Python.
The reason this blog exists is more personal. Learning this field can feel lonely. Scattered docs, half-finished tutorials, and late nights spent piecing things together. Every time an idea finally clicked for me, I wished someone had explained it clearly the first time. AnupTechTips is my attempt to be that person for you.
So I write the way I’d explain something to a friend: open about the trade-offs, clear about the mistakes I’ve made, and never pretending the hard parts are simple.
Why I do this
My goal is simple, and it means a lot to me. I want engineers, data practitioners, and curious builders to understand modern AI and scalable systems deeply, not just well enough to copy and paste. If a single post here saves you a late night, or makes a hard concept finally make sense, then it was worth writing.
Let’s connect
New posts land regularly, and I’d love for you to come along. Subscribe to the newsletter to get them in your inbox, or find me on LinkedIn, GitHub, Medium, and X. If something here helped you, confused you, or sparked a question, my contact page is always open, and I read every message.