Portfolio loading

Preparing portfolio 0%
GV Resume

Software Engineer / AI-ML Researcher / IIT Bhilai

Galaba Vamsi

I build scalable software and AI systems across full-stack products, wireless research, computer vision, GPU pipelines, and applied machine learning.

View Projects Resume GitHub LinkedIn
Pixel art portrait of Galaba Vamsi

About

Engineer with a research streak and a bias toward shipping.

I am a B.Tech Mechatronics student at IIT Bhilai, graduating in May 2027. My work sits at the intersection of software engineering, AI/ML, computer vision, wireless systems, shader programming, and distributed cloud infrastructure.

I like projects where the constraints are real: low budgets, limited hardware, messy data, and systems that need to work beyond a demo. My coursework includes data structures, object-oriented programming, deep learning for computer vision, AI/ML, digital image processing, and optimal control.

Experience

Selected Work

Research Intern - IIT Bhilai

May-Aug 2025

Advisor: Dr. Arzad Alam

  • Built a zero-budget open-source 6G RIS simulator for LoS/NLoS environments.
  • Emulated USRP-based workflows to reduce physical hardware and pre-deployment costs.
  • Designed and validated a two-layer RIS PCB workflow for RF energy harvesting research.
View RIS Simulator

Co-Founder - Human Slop

2026

Anti-AI social platform / YC Startup School

  • Built a React, PostgreSQL, and AWS platform focused on human-only social content.
  • Worked on hardware-bound biometric auth and real-time typing forensics.
  • Designed systems to block synthetic content and bot-like behavior.
Open Web App Play Store

Projects

Things I have built

Arista Wi-Fi Optimization System

Zero-budget Python simulator and AI-assisted RRM framework using Dual-CNNs, GNNs, and Safe RL. Ranked 5th at Inter IIT Tech Meet 14.0.

Python / Safe RL / GNN / Wi-Fi

Human Slop

Anti-AI social platform with biometric auth, typing forensics, PostgreSQL, AWS, and real-time synthetic content defense.

React / PostgreSQL / AWS

Web App

Electron-GNN

Two-tower hybrid GNN for predicting molecular absorption spectra from geometry, aimed at reducing expensive RT-TDDFT simulation needs.

PyTorch / GATv2 / Quantum ML

GitHub

Open-PyFX

GPU-accelerated Python and GLSL video effects suite for CRT emulation, ASCII edge detection, and hardware-accurate 3D LUTs.

Python / GLSL / GPU

GitHub

Lunar Crater Detection

Computer vision pipeline for extracting craters from Digital Elevation Models, with F1 0.839 and a CuPy/CUDA morphology backend.

OpenCV / CUDA / CuPy

GitHub

AntennaNet

Inverse EM design tooling with KD-tree nearest-neighbour anchoring, 144D search spaces, and Python-MEEP workflows on OpenMPI clusters.

MEEP / OpenMPI / Deep Learning

GitHub

Amazon ML Challenge

Top 3.5% solution over 20,000+ participants using Qwen 2-VL 7B and a ViT/RoBERTa two-tower model for multimodal price prediction.

Qwen / ViT / RoBERTa / ML

Serenity

Flutter mental wellness app with computer-vision-based real-time emotion detection, stress games, exercises, and doctor booking.

Flutter / Computer Vision / Mobile

GitHub

Blog

Build logs

One post per experiment: what I set out to test, the exact numbers I measured, and the parts that did not work.

PCA scatter plot of latent embeddings showing a smooth colour gradient by agent x-position.
  • Aug 4, 2026
  • 12 min read
  • Physical AI
  • JEPA

I trained a world model on my laptop. It learned physics and still could not push a block.

Experiment 01: a 1.9M-parameter JEPA world model on an 8 GB laptop GPU — seven minutes of training, emergent physical state, and a planner that scored exactly zero.

Read the write-up ->
All posts

Skills

Technical Focus

Software Engineering

Full-stack web and app development, GPU shader programming, Git, Linux, and UI/UX.

Languages

Python, C/C++, TypeScript, JavaScript, Kotlin, Swift, Dart, Go, GLSL, and CUDA.

Web, App, and Cloud

React.js, React Native with Expo, Node.js, Flutter, PostgreSQL, AWS, Supabase, and OpenMPI.

AI, ML, and Systems

PyTorch, GNNs, Safe RL, Transformers, OpenCV, CuPy, Wi-Fi PHY/MAC modeling, SDR/USRP, IEEE 802.11k/v, and PCB design.

Achievements

Highlights

1st PlaceIIT Bhilai Web Development Competition
YC Startup SchoolAccepted with Human Slop
5th RankArista Networks Wi-Fi Challenge, Inter IIT Tech Meet 14.0
Top 3.5%Amazon ML Challenge, 700 / 20,000+ participants
Top 1%JEE Advanced
Top 10 FinalistToyota Hackathon
Academics95% in Class 12 and 94% in Class 10

Contact

Open to software engineering and AI/ML research roles.

If your team needs someone comfortable moving between product engineering, ML systems, research prototypes, and low-level performance work, I would be happy to talk.

Email Me LinkedIn GitHub Portfolio Resume

galabavamsi12@gmail.com / +91 7483047846