What I've Built
Projects
A collection of things I've shipped — from AI experiments and open-source tools to weekend rabbit holes that got out of hand.
LLM Inference Engine in Rust
A Rust-based LLM inference engine focused on efficient model execution, token generation, and low-level systems performance.
Namaste Screen
🚀 Introducing NamasteScreen: The Launcher That Fights Phone Addiction! 🇮🇳 🔥 Inspired by my conversation with Hon'ble Prime Minister of India Narendra Modi about Digital Well-Being
Minecraft AI
AI playing Minecraft with me to test Minecraft civilization ideas and study how LLMs interact inside a human-based simulation.
Simulating Human Civilisation in Minecraft
Using LLM's and Minecraft we can create character's which are more than ncps
Decentralized Mesh Model Training
A lightweight framework that allows multiple Android phones to collaboratively train a model without centralizing data, with secure aggregation, partial model updates, and mesh-style communication.
Youtube Beyond 2x
Created a Extension that can help you finish the course 12x faster, ( FINISH not understand ) It can help you select the exact speed you want even 12.75x !! 😂😂😂 Here are the testimonials and FAQS, Source code below
N.E.T.H.R.A
N.E.T.H.R.A - Neural Emulated Terms Helping Resistance Assistance, A.I. that can help you in your office attedance through facial landmark detection
GopherQueue-System
GopherQueue-System is a Go-based asynchronous task processing system. It offers a simple API for queuing jobs and checking their status, using Gin for high performance. With concurrent workers and Docker support, GopherQueue-System efficiently handles time-consuming tasks, making it ideal for applications needing robust background job processing.
Multiverse Chat Engine
A branching conversation application for LLMs that stores chat as a tree, letting you fork conversations, regenerate responses, and explore multiple dialogue paths with Ollama.
Long Term Memory AI
A practical long-term memory system for LLM agents that models working, episodic, semantic, and procedural memory to make multi-turn AI behavior more context-aware and capable.
BlazFastChat
This chat app is made using Yew ( WebAssembly + Rust ) and Golang the fastest
SynthTextArt-Text-Guided-Image-Translation-GANS
Here I am just trying to implement this paper from scratch -> https://arxiv.org/pdf/1611.07004.pdf
Particle Physics System
This project is inspired js-ts particle tried to implement on of their most popular effect from scratch no libraries