AI/ML researcher and Computer Engineering student interested in building and understanding reliable, interpretable machine-learning systems.
My interests span explainable and trustworthy AI, language models, representation analysis, robustness, multimodal learning, and computer vision.
I am particularly interested in explainability, trustworthy machine learning, representation analysis, robustness, and the behavior of modern language and multimodal models.
I am an undergraduate Computer Engineering student at Sir Syed University of Engineering & Technology (SSUET), Pakistan, with research and project experience across machine learning, language models, computer vision, and explainable AI.
I am drawn to questions about what neural models learn internally, how those representations influence model behavior, and how we can evaluate systems more rigorously than by final-task accuracy alone. My interests include neuron-level interpretability, causal intervention, robustness, grounding, and multimodal learning.
Alongside research, I enjoy building practical AI systems and experimenting with tools such as PyTorch, Hugging Face, LangChain, OpenCV, and retrieval-augmented generation. I am interested in pursuing graduate research in AI and machine learning.
Selected research, teaching, and technical experience.
Research under Prof. Hassan Sajjad focused on neuron-level interpretation and causal intervention for explainable AI, including literature review, experimental design, representation analysis, intervention-based evaluation, and research prototyping.
Investigated machine-learning models for student performance prediction and customer segmentation using large-scale datasets. Designed and validated preprocessing strategies that improved model robustness and accuracy by 15%, and collaborated with engineers and data scientists to refine methodologies and interpret experimental results.
Selected twice as a volunteer Section Leader for Stanford's Code in Place, supporting learners through introductory programming concepts, problem solving, and hands-on Python practice.
Undergraduate coursework and projects spanning machine learning, deep learning, computer vision, software engineering, embedded systems, data structures, and computer networks.
A selected set of projects emphasizing experimentation, model behavior, and practical AI systems rather than a long catalogue of coursework.
Evaluated image classifiers under FGSM and PGD adversarial attacks, measuring degradation in predictive performance and exploring model robustness under perturbed inputs.
View repositoryBuilt a RAG assistant for YouTube videos using transcript retrieval, vector search, and language models to answer grounded questions through a lightweight Streamlit interface.
View repositoryDeveloped a document question-answering pipeline with retrieval-augmented generation and Gemini, focusing on context-aware responses and more grounded model outputs.
View repositoryBuilt a real-time sign-language recognition system using hand landmarks and classical machine learning, with text output, speech support, and gesture-based interaction.
Explore on GitHubSelected for a funded research internship at Dalhousie University, Canada.
Selected in 2025 and 2026 to mentor learners in introductory programming.
Completed all 9 of 9 puzzles with a perfect team score.
Ranked 103 out of 503 participants.
Received a 100% university merit scholarship as a semester position holder.
Machine Learning Specialization, LangChain Academy, and introductory ML on AWS.
Grouped by actual working areas, with no arbitrary percentage bars.
PyTorch · TensorFlow · scikit-learn · neural networks · CNNs · model evaluation · adversarial robustness
Transformers · Hugging Face · neuron-level analysis · causal intervention · RAG · LangChain · FAISS · Gemini
OpenCV · MediaPipe · image classification · visual recognition · ResNet · MobileNet · multimodal systems
Python · Git/GitHub · Linux · Docker · Streamlit · FastAPI · AWS · SQL · experimentation workflows
I am always happy to connect around explainable AI, trustworthy machine learning, language models, research internships, and graduate research opportunities.