Hi!
I’m Prabhsurat Singh — a software engineer
from New Delhi, India, with a Bachelor's degree in Computer
Science and a minor in Artificial Intelligence and Machine
Learning.
I spend most of my time building machine learning systems,
backend infrastructure, and developer tools. My primary interests
are Reinforcement Learning, LLM post-training, and Deep
Learning, where I work on everything from RL algorithms and
training pipelines to GPU programming and
model optimization.
Alongside ML research, I enjoy building distributed backend
systems and high-performance infrastructure using Python,
C++, PyTorch, CUDA, FastAPI, Docker, Redis, Kafka, and PostgreSQL.
I'm driven by a systems-first mindset—taking ambitious ideas
from research to production-ready software.
Latest Articles
Thoughts on RL, systems, development and more
Reinforcement Learning: Building GridWorld
A step-by-step guide to building a GridWorld environment from scratch with Gymnasium, state mappings, rewards, and rendering.
20 min read
Building a Local GenAI Application using TinyLlama
Lighweight AI applications that can run locally
15 min read
Hello World
My first blog post
5 min read
From Scratch
Implementations of research papers, books, and algorithms to understand them from first principles.
Featured Projects
A selection of my recent work
RL-From-Scratch
Implementation of various Reinforcement Learning algorithms and environments
rl-robotics-simulations
Implement Deep RL methods on MuJoCo environments, via command line interface