About Prashant Pandey

My Story

I did not begin with programming. In the early part of class 8, I did not even know the word itself. One day, while reading articles on a website, I came across the term “hacking” and started searching for it out of pure curiosity. That curiosity opened the door to the wider world of computers, systems, and software.

What started as curiosity slowly became commitment. I began with Python as my first language, learning mostly from YouTube and hands-on experimentation. Over time, I expanded into Java, C, HTML, CSS, and JavaScript. I built my first project, a simple e-commerce store, and that experience made one thing clear: I enjoy making ideas real.

I am not someone who only likes writing code. I like understanding systems from the inside. I want to know why a system behaves the way it does, what hidden rules shape it, and whether there is a better way to design it. That mindset has shaped how I learn, what I build, and how I think about technology.

What I Build

My work spans AI, computer vision, game engines, automation, and experimental software systems. I care about projects that are useful, intelligent, and technically interesting — not just projects that look impressive from the outside.

How I Think About Technology

My mindset as a builder is less about simply coding software and more about exploring unknown systems. I do not just enjoy making things work — I enjoy asking why they work, what hidden patterns exist, and whether the current way is the only way possible.

• Understand the problem first. Before I write code, I try to understand what is really causing the problem.
• Prefer meaningful simplicity. Strong systems should be powerful without becoming unreadable or fragile.
• Build to learn. I learn best by making real things, testing them, and improving them through experience.
• Question defaults. Standard methods are useful, but they are not always the best or only answer.
• Share knowledge openly. I believe ideas should be explored, questioned, and improved publicly.

Current Focus

My main direction is the intersection of advanced AI systems and next-generation interfaces. I am most interested in tools and models that feel intelligent, responsive, and closer to the way humans actually think and interact.

Featured Research Project

One of my most important projects is a human-like chess AI built with Python, PyTorch, and python-chess. The system is designed to learn from real human games, model openings, tactics, mistakes, and plans, and generate legal and natural moves that feel closer to human gameplay than brute-force engine output.

The workflow begins with Lichess human game datasets, then moves through PGN parsing and cleaning, board encoding, a lightweight transformer-based policy model, selective human-style search, and evaluation based on practical thinking instead of raw engine calculation. I first developed it on a small 8 GB Ryzen laptop, then scaled training and testing through Colab when needed.

This project matters to me because it represents the kind of research I love most: using careful design to model behavior, not just to maximize numerical performance.

Tools & Technologies

I choose tools based on the problem, not the trend. These are the languages and technologies I use most often or am actively growing with.

What Drives Me

What keeps me building is the feeling that there are still undiscovered patterns everywhere — a new theory, a better system, a different way to think, a mechanic nobody combined before, or an explanation nobody noticed. That possibility is deeply motivating to me.

My long-term goal is to become an independent researcher, AI creator, and experimental system builder. I want to create work that feels thoughtful, useful, and slightly ahead of what people expect from a single self-taught builder.

Journey