Portfolio

Deep Learning

Data Engineering

NLP

SMS Spam Classification

A comprehensive spam detection system implementing multiple text classification approaches including Bag of Words (BOW), TF-IDF Vectorization with Multinomial Naive Bayes, Word2Vec with Random Forest Classifier

Kindle Review Sentiment Analysis

Sentiment analysis on Amazon Kindle reviews using various NLP techniques: Implemented multiple text vectorization methods (BOW, TF-IDF, Word2Vec), built and compared different classification models, processed and analyzed the Kindle Reviews dataset.

Data Visualization

Call Centre Dashboard using Excel

Designed an interactive Excel dashboard using pivot tables, charts, and slicers. Tracked key metrics like call time, volume, and amount. Improved visibility into performance trends. Strengthened my Excel visualization skills.

Global Superstore Sales Dashboard

Built a Power BI dashboard to analyze global sales by region, category, and time. Included dynamic filters, forecasts, and decomposition trees. Enabled trend discovery and strategic comparison.

Power BI Pokedex

Created a Power BI dashboard inspired by Pokémon’s Pokedex. Integrated external data to build interactive profiles and visuals. Designed a fun, game-style user experience.

HR Analytics Dashboard

Built a Power BI dashboard to visualize headcount, salaries, leave data, and demographics. Used DAX and Power Query for dynamic insights. Supports strategic HR decisions through clear, interactive visuals.

8 Week SQL Challenge

Danny’s Diner SQL Case Study

Analyzed customer behavior and sales for a Japanese diner using PostgreSQL. Tackled real-world questions with CTEs and subqueries. Revealed purchase patterns and visit frequency.

Clique Bait SQL Case Study

Used SQL to evaluate ad performance and user behavior for an online seafood store. Found a 23% lift in purchases from ad engagement. Identified top campaigns and provided optimization strategies.

Machine Learning

HR Insights at Salifort Motors

Analyzed HR data to understand attrition drivers at Salifort Motors. Used EDA and ML (Random Forest, Decision Tree) for prediction. Achieved 96.2% accuracy, 93.8% AUC. Delivered actionable retention insights.

Student Performance Prediction

Predicted student academic outcomes using demographic and educational data. Applied EDA, preprocessing, and classification models. Helped identify at-risk students early.

Data Analytics

Walmart Sales Analysis using SQL

Explored Walmart transaction data to uncover sales trends, top products, and customer segments. Highlighted category performance and optimization areas. Used SQL CTEs for efficient querying.