I enjoy building things from first principles.
While my day job involves building large-scale web applications using React, Angular and TypeScript, most of my free time goes into exploring how software works beneath the abstraction layers.
That curiosity has led me to build:
- 🎮 A custom 3D renderer from scratch using WebGL
- ⚙️ Physics engine components (GJK, EPA, XPBD)
- 🧠 Neural network experiments in C
- 🤖 AI-powered applications using LLMs and RAG
- 🖥️ Operating system, compiler, emulator and systems programming projects
Frontend Engineer @ Darwinbox
Building enterprise-scale HR and Payroll software used by organizations across the globe.
Work includes:
- React & Angular applications
- TypeScript architecture
- Design systems & component libraries
- Performance optimization
- AI-powered workflow automation
- Full-stack feature development
Building rendering and simulation systems from scratch.
- Software Rendering
- WebGL
- Graphics Pipelines
- GLSL Shaders
- Physics Engines
- Collision Detection
- Spatial Data Structures
Interested in understanding how computers actually work.
- C++
- Linux
- Memory Management
- Networking
- Virtualization
- QEMU
- Operating Systems
Applying AI to real products instead of toy demos.
- LLM Integrations
- RAG Systems
- Prompt Engineering
- AI Agents
- Retrieval Systems
A lightweight renderer written from scratch without relying on Three.js or external graphics libraries.
🔗 https://sidx2.github.io/3d-renderer/
Experiments with rigid body physics, collision detection and constraint solvers.
🔗 https://sidx2.github.io/phyzzix/
An AI-powered mock visa interview system capable of generating contextual interview questions and evaluating responses in real time.
A collection of projects, experiments and engineering rabbit holes.
I enjoy solving hard problems.
🏆 LeetCode Knight
📈 Top 10% Global Ranking
🔗 https://leetcode.com/u/DevilsAlgorithm/
Areas of interest:
- Dynamic Programming
- Graph Algorithms
- Trees
- System Design
- Computational Geometry
TypeScript • JavaScript • Python • C++ • GLSL • SQL
React • Angular • StencilJS • HTML • CSS
Node.js • FastAPI • Express
PostgreSQL • MongoDB • Redis • Elasticsearch
Docker • AWS • RabbitMQ • Microservices
WebGL • Vulkan (Learning) • QEMU • Linux
The best way to understand a system is to build one.
That's why I enjoy creating things from scratch—even when a library already exists.
Whether it's a renderer, a physics engine, a neural network, or a web application, I like understanding the underlying mechanics rather than treating them as black boxes.
🌐 Website: https://sidx.in](https://sidx2.github.io/sidx/
💼 LinkedIn: https://linkedin.com/in/siddheshpardeshi
🏆 LeetCode: https://leetcode.com/u/DevilsAlgorithm/
🎮 Renderer: https://sidx2.github.io/3d-renderer/
⚛️ Physics Playground: https://sidx2.github.io/phyzzix/
"Build. Break. Understand. Repeat."
si
