Business Intelligence
Business Analyst-University Club
Tech Stack: Python, SQL, Google Sheets, Hypothesis testing, Regression, MySQL, Matplotlib, Seaborn, Tableau
Conducted comprehensive analysis of club attendance data by collecting and transformation using ETL. Reported the trends and strategies which resulted in 10% increase in event attendance and 30% increase in overall student engagement.
Campaign Analytics
Tech stack: Python (Pandas, NumPy, NLTK, Spacy, regex), Excel, Hugging Face, Data Patterning, Matplotlib, Tableau
Campaign analytics is where BI analysts continuously monitor and adjust their strategies. They can iterate and refine campaigns by analysing performance metrics, ensuring they remain relevant and effective.
Jaguar didn’t receive great response for its rebranding video. When analysed the gap, the result was interesting. Here you go for the results,
Recommendation using graphs
Tech Stack: Python, Scikit-learn, PyTorch, GNN, Matplotlib, Seaborn
Focused on addressing the cold start problem and tackling the challenges of modeling complex relationships between users and items.
Developed a personalized recommendation by utilising Collaborative Filtering based on user behaviour and preferences. Implemented Graph Attention Network attaining a precision rate @ 10 surpassing competitors by 60.7% enabling the system to give recommendations with few training data.
Comeback Nike - Case study on Nike’s Stocks
Tech Stack: NewsAPI, Yahoo Finance API, Python, Hypothesis testing, Time-series model(ARIMA), Matplotlib, Tableau
Risk Management is an operation where BI plays a critical role. Assessing and controlling threats to an organisation’s capital and earnings.
Nike ( in 2024) is facing a downward trend in the stock market, and here I have detailed the key events that could have caused it and reported the action plan to handle this impactful situation.
Machine Learning & AI
Sign Recognition System
Tech Stack: Python, Mediapipe, Scikit-learn, TensorFlow/Keras, NumPy, Pandas, Matplotlib, OpenCV, Computer Vision
Designed and developed the system to recognise the sign language to benefit the students with speech impairment and hearing loss
Achieved an 89% accuracy in the recognition model designed with media pipe
Research on Adversarial Attack
Tech Stack: Python, PyTorch, NumPy, Pandas, Matplotlib, Scikit-learn, Elastic Net for adversarial attacks, Transfer Learning, Research
Achieved 95% accuracy with an Inception ResNet model for MRI eye image classification
Implemented and monitored Elastic Net attacks, which reduced the accuracy to 7.5%; Programmed transfer learning to increase accuracy to 70% and fortified the model against vulnerabilities.
Data Analysis
Cyclist Company Case Study
Tech Stack: Python, ETL, EDA, Matplotlib
Employed the six stages of the Data Analysis process—Ask, Prepare, Process, Analyze, Share, and Act—to evaluate user engagement strategies for a Bicycle Company
Collected and cleaned the data from user activity logs, surveys, and digital media metrics to understand casual riders' behaviour. Used ETL pipelines to transform the data for exploratory and statistical analysis.
Crafted personalized marketing campaigns and strategies to improve conversion rates from casual riders to annual members.