Data Scientist

Technical Skills:

Internship Project by Boston Consulting Group Link

Conducted in-depth research to identify key factors contributing to customer churn.
Executed various statistical and quantitative analyses to uncover insights and patterns.
Developed and trained a Random Forest model to predict customer churn with high accuracy.
End Goal: Enhanced client retention strategies by providing actionable insights and predictive capabilities to reduce customer attrition and optimize business performance.

Customer Churn

Internship Project Link

Utilized Python to analyze a comprehensive COVID-19 dataset, employing statistical modeling and hypothesis testing to derive actionable business insights.
Aggregated and visualized data using pandas and Seaborn, producing clear and informative visualizations and a detailed professional report.
Developed a Classification Machine Learning model to categorize COVID-19 severity, enhancing the ability to predict and manage case outcomes effectively.
End Goal: Provided valuable insights and a robust predictive tool to aid in understanding and managing COVID-19 severity, supporting informed decision-making and public health responses.

Covid Study

Opensource Project Link

Conducted data cleaning and preprocessing on 27,000 blood sample images to prepare them for analysis.
Applied Convolutional Neural Networks (CNN) to classify blood samples as parasitic or non-infected.
Achieved a model performance with 99% test accuracy and 94% validation accuracy through rigorous training and evaluation.
End Goal: Developed a highly accurate diagnostic tool to improve malaria detection, aiding in early diagnosis and treatment, and contributing to more effective malaria control strategies.

Malaria Detection Algorithm

Opensource Project Link

Fine-tuned DenseNet121 and developed a custom CNN, achieving 91% accuracy with 95% recall for pneumonia detection.
Delivered balanced precision for reliable diagnostics, minimizing false positives and negatives.
End Goal: Created an integration-ready model to support healthcare professionals in accurate and efficient pneumonia diagnosis.

Pneumonia Detection Algorithm

Opensource Project Link

Utilized MySQL to extract and store data from websites.
Built and trained an NLP model to classify websites based on their metadata.
Developed a web scraper using Selenium and BeautifulSoup for automated data extraction.
End Goal: Enhanced data management and classification efficiency by automating website data extraction and analysis.

AI Scrapper

Opensource Project Link

Developed a CNN model to identify 120 different dog breeds.
Trained on 10,000 images and tested on an additional 10,000 images.
Utilized tools for image data visualization and processing.
Achieved 98% test accuracy and 67% validation accuracy, followed by further fine-tuning.
End Goal: Created a high-accuracy model for dog breed identification, with potential for further refinement and real-world application.

Dog identifier model

Acquired practical skills in implementing various machine learning algorithms to meet diverse business requirements.
Completed assignments using more than six industry-relevant datasets, applying ML techniques to solve real-world problems.
Certificates : Link
Certificate 2 : Link

Roles & Responsibilities : Analyzing data, Building Predictive model
Goals : Minimizing client’s customer churns, optimizing business decisions.
Finalization : Successfully trained a model to predict customer churns, reasons for churn is addressed and optimized.
Certificate : Link

Exposure to various professional settings is various companies.
Delivering creatives, UI/UX, motion graphics and static graphics for websites & social media.
Proficient in Adobe creative suite, Adobe After effects, Adobe Photoshop & Adobe Premiere pro.
Promotional Content creation for business needs.
Portfolio Link : Link