Data Scientist with a strong academic foundation in Artificial Intelligence, skilled in building end-to-end ML pipelines, statistical analysis, and data-driven storytelling. Passionate about turning complex data into intelligent decisions.
Data Scientist
I'm Muhammad Umer, a Data Scientist pursuing a BS in Artificial Intelligence at the University of Sialkot. I specialise in building end-to-end data science pipelines — from raw data collection and preprocessing through to machine learning model development, evaluation, and deployment.
My core toolkit covers Python, Pandas, NumPy, Scikit-Learn, TensorFlow, Keras, Matplotlib, Seaborn, Plotly, Power BI, SQL, and Excel. I work across supervised and unsupervised learning, deep learning architectures, NLP pipelines, and computer vision tasks.
I've already shipped a live AI-powered web application — an AI Data Analyst Agent that lets users upload any dataset and query insights in plain English — and I'm building more production-grade ML projects to demonstrate real capability.
Shipped an AI Data Analyst Agent on Render — users upload any CSV and query it with natural language via Gemini AI.
Experienced in the full data science workflow — EDA, feature engineering, model training, evaluation, and deployment.
Semester 7 at University of Sialkot — coursework covering ML, Deep Learning, NLP, Computer Vision, and AI Systems.
University projects in CNN optimisation, RNN vs LSTM action recognition, BERT-based NLP, and RAG system design.
Built through academic projects, independent study, and shipped production applications.
A full-stack web application that lets users upload any CSV dataset and query it using plain English. Powered by Google Gemini AI, Flask, Pandas, and Plotly — it generates instant charts, statistics, and insights from natural language questions.
Supervised learning model trained and evaluated on structured data with full performance reporting.
Visual analytics dashboard built in Power BI or Python to communicate complex data clearly.
Deep learning model built from scratch using TensorFlow/Keras for a classification task.
Natural language processing pipeline for text classification or sentiment analysis on real data.
Full end-to-end AI/ML solution combining data engineering, model building, and deployment.
Data science solutions delivered with rigour, transparency, and a focus on real measurable outcomes.
Remove duplicates, fix errors, standardize formats and prepare raw datasets for reliable analysis.
Transform numbers into clear, compelling charts and graphs that communicate your data story.
Deep analysis of your datasets to uncover trends, patterns, and hidden opportunities for growth.
Interactive, real-time dashboards that give you a clear overview of your key metrics at a glance.
Professional Excel dashboards with pivot tables, formulas, and automated reports for your business.
Custom, publication-ready plots in Python (Matplotlib, Seaborn, Plotly) for presentations or reports.
Extract, collect, and structure valuable information from large datasets and multiple data sources.
Restructure, merge, reshape and format datasets into clean, analysis-ready files quickly and accurately.
Have a data or ML problem?
Let's solve it.
I'm open to data science projects, ML collaborations, research work, and internship opportunities. If you have a dataset, a problem, or an idea — let's talk.