Welcome to my portfolio

Shanker Lal Meghwar

Machine Learning & Data Science

Budapest, Hungarymeghwarshanker24@gmail.com

About Me

I'm a Computer Science student at the University of Pécs, Hungary, specializing in Machine Learning and Data Science.

Currently working on my final-year thesis focused on diabetes prediction using ML models on imbalanced datasets—tackling real-world clinical challenges with data-driven solutions.

I combine technical skills in Python, Java, and C with expertise in XGBoost, Random Forest, and scikit-learn to build impactful machine learning solutions.

ML Models Built

350K+

Datasets Processed

0.92

Best AUC Score

200%+

Recall Improvement

Experience

Internship: Embedded Systems & Automation

Computing & Education Technology Centre, University of Pécs

Feb 2026 – Apr 2026

Pécs, Hungary

SFTP Debugging: Fixed directory-handling bug between MySecureShell and Paramiko (root mapping issue). Reproduced locally, patched path logic, documented solution.
IoT Hardware Integration: Built event-driven system with ESP8266 + physical switches triggering HTTP requests over Wi-Fi. Handled sensor disconnection edge cases.
Automation Pipeline: Designed file-monitoring system (detect → process → send → delete) implementing complete end-to-end workflow in Linux environment.

Operations Team Member & Social Work Intern

Foundation Fighting Poverty & The Blood Heroes

09/2022 – 07/2023

Karachi, Pakistan

Designed and led volunteer training programs, improving team efficiency by 25%
Led fundraising campaigns, raising over PKR 100,000 to aid marginalized communities
Expanded outreach by 20% through community engagement initiatives
Organized three blood drives, surpassing donation targets by 15% and supporting 100+ patients

Education

Bachelor of Science, Computer Science (Data Science Track)

University of Pécs

Sep 2023 – Jun 2026

Budapest, Hungary

Deep LearningAIData StructuresDatabase ManagementMachine LearningPythonJavaLinuxStatistics

Projects

Final Year Thesis

Diabetes Prediction Using Machine Learning

Developed ML models for clinical diabetes prediction on large-scale, imbalanced datasets.

Trained XGBoost, Random Forest, Logistic Regression on ~100K clinical & ~254K BRFSS records
Tackled severe class imbalance (46:1 ratio) → achieved >200% recall improvement
Key results: Clinical AUC 0.92, Recall 0.85 | BRFSS Recall 0.88 (from 0.28)
Methods: ADASYN resampling, class weighting, hyperparameter tuning, threshold optimization

Professional Experience

Internship: Embedded Systems & Automation

Computing & Education Technology Centre, University of Pécs

SFTP Debugging: Fixed directory-handling bug between MySecureShell and Paramiko (root mapping issue)
IoT Hardware: Built event-driven system with ESP8266 + physical switches triggering HTTP requests
Automation Pipeline: Designed file-monitoring system (detect → process → send → delete)

Academic Project

University Administration System

Built a Java and Spring Boot-based university system for efficient student record and workflow management.

Integrated MySQL for secure data handling
Employed SQL queries for analysis and reporting
Documented APIs with Swagger for seamless integration

Skills

Programming

PythonJavaC

Machine Learning

XGBoostRandom ForestLogistic RegressionScikit-learn

Data Analysis

NumPyPandasSQLJupyter Notebook

Tools & Technologies

GitLinuxMySQLESP8266SFTP/SSH

Frameworks

Spring BootPyTorchTensorFlow

Let's Connect

I'm always open to discussing new opportunities, projects, or just having a conversation.

Email GitHub LinkedIn

Available for opportunities