M.Tech @ NIT Kurukshetra · ML Researcher · Open to Opportunities

Yuvraj Chulpar

ML Engineer & Researcher Federated Learning Specialist LLM Fine-tuning Engineer GPU Computing Enthusiast
const engineer = {
  name: "Yuvraj Chulpar",
  location: "NIT Kurukshetra, India",
  specialization: ["Federated Learning", "LLMs", "RAG"],
  kaggle: "Competitor · Nemotron Rank #388",
  status: "building_cool_stuff"
};

01 · About

Who I Am

I'm an ML Engineer and Researcher pursuing M.Tech in Computer Engineering at NIT Kurukshetra, specializing in Federated Learning, GPU-accelerated Computer Vision, and LLM fine-tuning.

My research focuses on Non-IID data challenges in Federated Learning, where I developed Clust-PSI-PFL — a novel clustering framework using Population Stability Index that reduces communication overhead by 40% and improves model accuracy by 15% on skewed datasets.

I'm currently extending this work with formal security mechanisms — integrating Differential Privacy and Byzantine-robust aggregation into the FL pipeline. I also build production-grade GenAI systems including RAG pipelines with Mistral-7B fine-tuning.

Outside research, I lead campus placement operations as TPO Coordinator at NIT Kurukshetra, coordinating 50+ companies and facilitating placements for 150+ students annually.

500+
DSA Problems Solved
40%
Comm. Overhead Reduced
98%
CNN Model Accuracy
150+
Students Placed Annually

02 · Research

Publications & Research

Under Review · Actively Extending

Clust-PSI-PFL: A Population Stability Index Approach for Clustered Non-IID Personalized Federated Learning

Yuvraj Chulpar et al. · NIT Kurukshetra · 2025

We propose a novel federated learning framework that leverages Population Stability Index (PSI) as a lightweight client clustering metric to address Non-IID data heterogeneity. By grouping clients with similar data distributions via K-means++ on PSI features, our method achieves 15% accuracy improvement on CIFAR-10/FEMNIST while reducing communication overhead by 40% compared to gradient-based clustering methods. Benchmarked against FedAvg, FedProx, and Per-FedAvg — outperforming all on Non-IID partitions.

Federated Learning Non-IID Data Population Stability Index Personalized FL K-means++ PyTorch CUDA
Skew Generalization
Extending PSI applicability beyond label skew to attribute skew and quantity skew scenarios
Privacy Mechanisms
Integrating Differential Privacy (DP) and Secure Multi-Party Computation (MPC) into the FL pipeline
Security Extension
Designing threat models for Byzantine-robust aggregation and adversarial client detection

03 · Projects

What I've Built

PROJECT_01

Personalized Federated Learning Framework (Clust-PSI-PFL)

Novel clustering-based FL system tackling Non-IID data challenges across distributed GPU-accelerated networks using Population Stability Index as a lightweight clustering metric.

−40% comm. overhead +15% accuracy Multi-node training
Python PyTorch CUDA Flower PySyft
View on GitHub →
PROJECT_02

RAG-Based Intelligent Document QA System

End-to-end RAG pipeline with FAISS dense + BM25 sparse hybrid retrieval using Reciprocal Rank Fusion. Fine-tuned Mistral-7B with QLoRA for domain-specific question answering over research papers.

+22% accuracy +18% recall@5 <200ms latency
Python LangChain Mistral-7B FAISS QLoRA Azure
View on GitHub →
PROJECT_03

Plant Leaf Disease Detection & Fertilizer Recommendation

GPU-accelerated CNN achieving 98% accuracy across 30+ disease categories. Automated fertilizer recommendation module mapping detected diseases to specific treatments.

98% accuracy 30+ categories 5x data expansion
Python TensorFlow Keras CNN OpenCV
View on GitHub →

04 · Skills

Tech Stack

AI & Machine Learning
Deep Learning (CNNs, Transformers)92%
Federated Learning95%
LLM Fine-tuning (QLoRA/LoRA)85%
RAG Pipelines88%
Computer Vision82%
Engineering & Tools
Python95%
PyTorch / CUDA88%
C/C++75%
Cloud (Azure / AWS)72%
Docker / Linux80%
PyTorch
TensorFlow
HuggingFace
LangChain
FAISS
Flower/PySyft
CUDA
Scikit-learn
OpenCV
Pandas/NumPy
Docker
Azure

05 · Education

Academic Background

M.Tech in Computer Engineering
2025 – 2027 (Expected)
National Institute of Technology, Kurukshetra · Haryana, India
  • Specializing in Machine Learning, Deep Learning, Distributed Systems, and GPU Computing
  • Research focus: Federated Learning, Non-IID data heterogeneity, Privacy-preserving ML
  • Relevant Coursework: Advanced Machine Learning, Computer Vision, Data Mining, Computer Architecture
B.Tech in Computer Science and Engineering
2020 – 2024
Government College of Engineering, Chh. Sambhajinagar · Maharashtra, India
  • CGPA: 8.02 / 10.0
  • Head of UI/UX Design at CATALYST Technical Club — led 10+ member team for 3 major technical events
  • Built foundational expertise in Data Structures, Algorithms, OOP, Operating Systems, and Computer Networks

06 · Experience

Academic Experience

Training & Placement Coordinator
2026 – Present
National Institute of Technology, Kurukshetra
  • Coordinating campus recruitment drives scaling from 10+ to 50+ companies, facilitating placement for 150+ students annually
  • Organizing pre-placement talks, workshops, and mock interview sessions, improving placement readiness by 30%
  • Managing a team of 33 student volunteers for end-to-end placement activities
  • Streamlining recruitment pipeline — scheduling, logistics, and feedback for 30+ interview panels per semester
Head of UI/UX Design
Jul 2022 – May 2024
CATALYST Technical Club · GCoE Sambhajinagar
  • Led a team of 10+ designers and developers to deliver web interfaces and branding assets for 3 major technical events
  • Promoted from Designer to Head within 6 months based on design quality and leadership performance
  • Prototyped and shipped responsive interfaces with modern UI/UX principles

07 · Achievements

Recognition

Kaggle Competitor · Nemotron Rank #388
Actively competing in the NVIDIA Nemotron Model Reasoning Challenge — applying QLoRA fine-tuning and synthetic data generation on Blackwell GPU infrastructure. Previously participated in Playground Series Heart Disease prediction with 4370 teams.
🎓
M.Tech @ NIT Kurukshetra
Pursuing M.Tech in Computer Engineering at National Institute of Technology, Kurukshetra — one of India's premier technical institutions — specializing in ML and distributed systems.
📄
Research Publication (Under Review)
Clust-PSI-PFL paper under review — novel FL framework achieving 40% communication overhead reduction and 15% accuracy improvement on Non-IID benchmarks.
🔗
Open Source · RAG Pipeline
Publicly released RAG-Document-QA system on GitHub — end-to-end Mistral-7B fine-tuning with hybrid retrieval. Available at github.com/chulparyuvraj/rag-document-qa
🎯
NVIDIA Nemotron Competition
Actively competing in the NVIDIA Nemotron Model Reasoning Challenge on Kaggle, applying fine-tuning techniques to improve reasoning accuracy on G4 infrastructure.
👥
TPO Coordinator Impact
Facilitated placements for 150+ students annually at NIT Kurukshetra, managing 33 volunteers and 50+ recruiting companies with 30% improvement in placement readiness.