Sanober Khan

AI & Machine Learning Specialist and Data Science Analyst with strong expertise in predictive modeling, statistical analysis, and enterprise IT project management. Skilled in Python, SQL, PL/SQL, Oracle APEX, machine learning frameworks, and data visualization, with a proven track record of turning complex business problems into actionable, data-driven solutions.

Successfully led multi-stakeholder IT portfolios at Bank AL Habib, overseeing 213 projects across digital and core banking while optimizing delivery pipelines, designing dashboards, and improving project throughput by 15–20%. Experienced in building production-grade database systems, developing regression and classification models (Random Forest, XGBoost), performing clustering (K-Means), and creating interactive BI dashboards for strategic decision-making.

Recognized for leadership, analytical excellence, and academic achievement (Dean’s List, top 10 in batch), with a strong ability to deliver scalable end-to-end solutions that bridge business strategy, AI/ML, and operational efficiency.

• Portfolio Coordination Lead: Led cross-functional coordination for a 213-project IT portfolio, aligning developers, testers, BAs, and business heads to meet regulatory and delivery timelines at Bank AL HABIB KARACHI
• Data-Driven Decision Leadership: Guided senior management using analytics-driven dashboards & insights to optimize resource allocation project throughput at Bank Al Habib
• Made into Dean's List at IBA
• Distinction in Numerical Analysis at IBA, Karachi
• Distinction in Arabic Language-1 at IBA, Karachi
• Among top 10 students in the batch of 2020, IBA Karachi

1) Project Title: Full-Stack Enterprise Database Management System with Oracle APEX 

Tools/Technologies Used: Oracle Database 19c/21c, Oracle APEX, SQL, PL/SQL, ERD/Data Modeling (draw.io), DDL/DML, Views, Stored Procedures, Functions, Unit Testing, Interactive Reports/Grids, Faceted Search, Data Visualization (Charts), HTML/CSS, Version Control

Brief Description & Results:

• Designed and deployed a production-grade enterprise database system end-to-end, modeling 13 normalized entities to support manufacturing, retail, and corporate operations.
• Built a robust data layer using optimized DDL, constraints, indexes, and validation rules, ensuring full referential integrity across 2,700+ records.
• Delivered actionable analytics by creating 7 SQL views and 20+ complex queries (joins, subqueries, aggregations) to identify sales trends, salary insights, and inventory risks.
• Automated business logic with PL/SQL, developing 5 procedures with validation and testing to standardize HR, inventory, and order-processing workflows.
• Implemented an interactive Oracle APEX application with 15+ data-driven pages, dashboards, and visualizations, enabling self-service reporting for non-technical users.

2) Project Title: Supermarket Sales Data Analysis & Predictive Modeling

Tools/Technologies Used: Python (Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, Statsmodels), Jupyter Notebook, Statistical Analysis (OLS Regression, Hypothesis Testing), Data Visualization, Clustering (K-Means), Kaggle Datasets

Brief Description & Results:

• Conducted end-to-end BI and ML analysis on a supermarket dataset (9,994 transactions), solving 8 business problems across demand forecasting, pricing, inventory, and supply chain.
• Built an analytics pipeline using statistics, trend analysis, and visualizations, identifying West region (55% revenue), 25% sales growth, and top-profit category (Eggs, Meat & Fish).
• Developed and validated regression models (simple & multiple), proving sales as the primary profit driver (R² = 0.368) and statistically confirming discounts had no significant impact (p = 0.262).
• Implemented a region-aware profit prediction model by encoding categorical variables to forecast profits based on sales and region for planning and budgeting.
• Applied K-Means clustering to segment product categories, enabling targeted marketing, optimized product placement, and cross-selling strategies.
• Delivered executive-level recommendations translating insights into actions, including budget reallocation, inventory optimization, and avoiding deep discounts on low-margin products.

3) Project Title: Advanced Statistical Analysis & Hypothesis Testing in Business & Cybersecurity

Tools/Technologies Used: SPSS, Statistical Hypothesis Testing, Pearson Correlation, Linear Regression, Chi-Square Test of Independence, ANOVA, Post-Hoc Analysis (Tukey's HSD), Shapiro-Wilk Test, Levene's Test, Data Visualization (Box Plots, Scatter Plots, Bar Charts)

Brief Description & Results:

• Result: Conducted rigorous statistical validation across three business scenarios, correcting flawed analyses to prove a strong correlation (r=0.98) between project delays and profit loss and identifying significant differences in cybersecurity response times across attack types (ANOVA, p<.001).
• Result: Audited and corrected critical statistical models, building a predictive regression (R²=0.96) that quantifies profit impact from delays and performing ANOVA to reveal key vulnerabilities in cybersecurity response protocols.

4) Project Title: Autism Spectrum Disorder Screening ML System

Tools/Technologies Used: Python, Scikit-learn, XGBoost, Pandas, Matplotlib, SHAP, LIME, Joblib

Brief Description & Results: Developed an end-to-end machine learning pipeline for early ASD detection, comparing 4 classification models with comprehensive validation and explainable AI techniques.
Achieved 92.4% accuracy and 0.91 F1-score with Random Forest classifier, reducing false negatives by 18% compared to baseline, with full interpretability using SHAP/LIME for clinical transparency.

5) Project Title: Amazon Product Review Sentiment Analysis with Comparative ML Evaluation

Tools/Technologies Used: Scikit-learn, TensorFlow/Keras, NLTK, spacy, TextBlob , Pandas, NumPy Matplotlib, Seaborn,

Brief Description & Results: Created a production-ready sentiment API with 92.8% accuracy that transformed Amazon review data into actionable business insights.

6) Project Title: World Master Chef – Full-Stack Social Media Platform for Food Enthusiasts

URL: https://mayar.abertay.ac.uk/~2411535/index.php#

Tools/Technologies Used: PHP (OOP), MySQL, JavaScript (ES6), HTML5, CSS3, Bootstrap 5, PDO, bcrypt, LocalStorage, AJAX, Draw.io, WebAIM

Brief Description & Results:

• Developed a full-stack culinary social platform with 9+ interactive modules including a Social Feed, Community Hub, Battle Arena, and E-commerce Menu.
• Built a secure backend with OOP PHP, featuring bcrypt authentication, session management, and PDO prepared statements for SQL injection protection.
• Implemented an SPA front-end using AJAX/JavaScript for seamless, reload-free navigation across user feeds, profiles, and tournament battles.
• Designed a responsive, mobile-first UI with Bootstrap 5 and custom media queries, validated for WCAG AA/AAA accessibility compliance.
• Integrated advanced features like simulated live chat, a persistent LocalStorage shopping cart, dynamic user search, and an interactive post/comment/like system.
• Deployed a fully functional production application showcasing end-to-end development, from security and gamification to professional UX/UI design.