Hello! I Am Param Pandya
Hello! I Am Param Pandya
An Engineer who
Builds intelligent
systems using AI & Data
Turning research ideas into scalable, real-world machine learning solutions.
I'm a |
Currently focused on AI/ML research and software engineering.
An AI/ML researcher and software engineer with a strong foundation in machine learning, deep learning, and NLP. I focus on building intelligent, scalable systems and applying research to real-world problems in healthcare and data-driven applications.
Experience
Research Intern
IIT Jammu • May - 2023 To July - 2023
Worked on machine learning and deep learning models with a focus on research-oriented problem solving and experimentation.
Research Intern
IIT Indore • May - 2022 to July - 2022
Conducted research in AI/ML with emphasis on data analysis, model evaluation, and academic research workflows.
Data Analyst (Machine Learning) Intern
upskillz.in • Aug - 2022 to Oct - 2022
Applied machine learning techniques for data analysis, preprocessing, and predictive modeling on real-world datasets.
Research Author
IEEE Conference • 2024
Published a research paper on efficient deepfake detection using AI, focusing on model robustness and performance evaluation.
I am an AI/ML researcher and software engineer with a strong academic background and a research-driven approach to problem solving.
My work focuses on deep learning, NLP, generative AI, and medical image analysis, with an emphasis on building robust and clinically meaningful AI systems.
Featured Research Project
Efficient Deepfake Detection using AI(2024)
An AI-driven deepfake detection system developed and published in an IEEE conference. The project focuses on robust feature extraction, deep learning architectures, and performance evaluation to detect manipulated media effectively.
Featured Research Project
PneuSTACK: Multi-Class Pneumonia Detection using Deep Learning
A stacking-based deep learning framework for automated pneumonia detection from chest X-ray images. The project introduces a novel perspective-distortion augmentation strategy and combines multiple pretrained CNNs with meta-learners (Logistic Regression, Linear Regression, and XGBoost) to improve robustness and diagnostic accuracy in both binary and multi-class settings. Vision Transformers (ViT) are also evaluated for comparative analysis.
