I started my blogging journey because I was tired of reading AI explanations that felt like they were written for people who already knew everything.

I'm Rushi Prajapati. I'm pursuing my Master's in Data Science and Quantitative Economics at Fordham University in New York, where I also work as a Graduate Research Assistant researching bias detection in Vision-Language Models. Before grad school I spent two years as an Associate AI Engineer at Sahana System Limited, building production AI systems for government, defense, and enterprise clients across five industry verticals. I've published peer-reviewed research in Springer Nature on reinforcement learning and in Elsevier on transformer architectures for image captioning. I won the AWS DeepRacer League in Gujarat with the fastest lap time of 11.39 seconds across 123 participants. I serve as Community Outreach Manager for Google Developer Group NYC, where I organize Build with AI Hackathons at Columbia, NYU, and Yale and MC events for hundreds of people in the New York tech community.

That's the resume version. Here's the real version.

I grew up genuinely fascinated by how machines learn things. Not the theoretical version of that fascination but the practical one where you stay up late trying to understand why your model isn't converging and you learn more in that frustrated hour than you did in a week of lectures. That curiosity is what led me to computer vision projects, to reinforcement learning research, to building platforms that let non-technical people access AI without needing an engineering degree. It's also what led me to start writing.

The Simplifying Series exists because the gap between what researchers understand and what most people understand about AI is enormous and mostly unnecessary. The concepts aren't as hard as the papers make them sound. The math isn't as intimidating as the notation suggests. And the implications of these technologies for how we live and work are too important to be locked behind academic jargon that most people reasonably give up trying to parse.

So that's what I do here. I take things that feel complicated and I find the version of them that actually makes sense to a curious person who isn't a specialist. Sometimes that's a deep dive into how a specific model architecture works. Sometimes it's a practical guide to building something. Sometimes it's an honest conversation about what AI can and genuinely cannot do.

If you're curious about where AI is going and how it actually works under the hood, you're in the right place. Follow along and let's figure it out together.