header

Hi, I'm Abdullah
I craft intelligent solutions

Introduction

Overview

Data Scientist with a passion for turning data into valuable insights. I use data analysis and machine learning to create solutions that help businesses make better decisions.

ML/AI Engineer

ML/AI Engineer

NLP Engineer

NLP Engineer

MLOps Engineer

MLOps Engineer

Data Visualization Expert

Data Visualization Expert


My Journey

Work Experience

Undergraduate Research Assistant - LLMs

FAST NUCES Lahore

Feb 2025 - Present

  • Fine-tuned pre-trained LLMs for domain-specific datasets, improving performance on various NLP benchmarks.
  • Collaborating with faculty members on research papers addressing the challenges and potential of LLMs in real-world applications.
  • Evaluating LLMs for bias detection, ethics, and fairness in NLP models.

Undergraduate Teacher Assistant

FAST NUCES Lahore

Jan 2025 - Present

  • Teaching Assistant for Software Engineering (CS-2004). Guided students through complex programming concepts, and assisted in debugging and problem-solving. Evaluated assignments, held office hours, and collaborated with faculty to improve student learning experiences.

Data Science Intern

Bytewise Limited

May 2024 - Sep 2024

  • Implemented machine learning models (regression, classification, clustering, and neural networks), achieving 90% predictive accuracy on real-world datasets.
  • Conducted data preprocessing and feature engineering to enhance model performance and insights delivery.
  • Fine-tuned deep learning models (Resnet150, EfficientNet-B)etc on real world datasets.

My Work

Projects

Explore a showcase of my diverse projects, highlighting my expertise in machine learning, AI-driven solutions, and advanced data analytics. Each project reflects my passion for solving real-world challenges through innovative and impactful digital experiences, using cutting-edge technologies and methodologies.

portfolio

CryoET Protein Complex Annotation & Classification Model

Healthcare

Developed a deep learning model to classify five protein classes (ribosome, virus-like particles, apo-ferritin, thyroglobulin, and β-galactosidase) in cryoET datasets using ResNet-50 and DenseNet-121 for feature extraction, and U-Net for segmentation.

#Tensorflow #OpenCV #Numpy
portfolio

Voice to Text Summarizer

AI & Machine learning

Fine-tuned HuBERT for voice-to-text conversion and BERT for summarization, enabling accurate transcription and concise summary generation from audio inputs. Achieving 92% accuracy

#Hubert #Bert #Pytorch
portfolio

DermaStratif - Multiclass Lesion Stratification and Diagnosis

Healthcare

Developed the DermaStratif model using EfficientNet-B0 with Adapter and LoRA fine-tuning, improving accuracy from 74% to 80%. Classified skin lesions into 8 categories for skin cancer diagnosis.

PyTorch #Resnet150 #Efficientnet B
portfolio

Nail Morphology Classification and Color Detection

Healthcare

Developed a real-time Nail Detection System using TensorFlow, OpenCV, and YOLO, achieving 92% accuracy in detecting nail shapes and colors from live webcam feeds. Designed a user-friendly interface for beauty technicians.

#YOLO #scikit-learn #Tensorflow
portfolio

Upwork Data Analysis & Budget/Skills Forecasting

Business & Consulting

Analyzed Upwork project data in Power BI to predict budgets, skills, and timelines. Developed and compared RNN, ANN, Random Forest, and Decision Tree models, with RNN achieving 89% accuracy.

#Plotly #Pandas #scikit-learn
portfolio

Recommendation Systems

AI & Machine Learning

Built a recipe recommendation system with Python, using content-based and collaborative filtering, NLP for ingredient extraction, and Scikit-learn for personalized suggestions.

#NLTK #Pandas #scikit-learn
portfolio

Nova - AI virtual assistant

Artificial Intelligence

Developed an AI virtual assistant using Python and OpenAI APIs for automated task handling and user interaction.

#OpenAI #Flask #scikit-learn
Your Portfolio

Muhammad Abdullah Naeem
Let's Talk!

An open invitation to connect, and exploring collaborative opportunities on my personal portfolio website.

Hover to Display Email