Portfolio - Manoj AA

Manoj AA

Data Aspirant

Resume

1

Python Project

2

Machine Learning Projects

About Me

Hi, I'm Manoj, passionate about Data Science. I love exploring data, uncovering insights, and building real-world solutions with the help of Machine Learning and Python.

One interesting fact about my journey is—I started applying Data Science concepts through hands-on projects like a Diabetes Prediction System using Machine Learning and a Real-Estate Price Prediction System which gave me the confidence to move from just learning to creating impact-driven solutions.

Though I’m a fresher, my project experience, combined with my background in programming (Data anlysis, Python, MySQL, Machine Learning), has helped me develop a strong foundation to analyze problems, build models, and deliver practical results.

Tools & Technologies I work with:
-Python
-SQL
-Machine Learning
-HTML, CSS, JavaScript

My Functional Areas:
-Data Analysis & Prediction
-Machine Learning Model Development
-Data Visualization

Key Skills

Python

Numpy

Pandas

Machine Learning

Data Science

Scikit-learn

My Projects

Expense Tracking System

Domain/Function: Software Development

Credit Card Launch

Domain/Function: Statistical Data Analysis

Health Insurance Premium Prediction

Domain/Function: Machine Learning

Credit Risk Prediction

Domain/Function: Machine Learning

My Experience

- Tech Stack: Python, Pandas, NumPy, Scikit-learn, Matplotlib
  Key Contributions:
  Worked on a Real Estate Price Prediction project using machine learning regression techniques.
  Performed data cleaning, EDA, and feature engineering on real-world housing data.
  Built and evaluated models to predict prices based on location, size, and BHK.
  Designed REST APIs using Flask to connect predictions with a user interface.
  Used Postman and CORS to test and ensure smooth API communication.

Awards & Certificate

Python: Beginner to Advanced For Data Professionals

SQL for Data Science

Math and Statistics For AI, Data Science

Master Machine Learning for Data Science & AI: Beginner to Advanced

Let's Connect

Feel free to get in touch with me. I am always open to discussing new projects, creative ideas or opportunities to be part of your visions.

Download Resume

Resume

Mail Me

[email protected]

https://www.linkedin.com/in/manoj-a-a-659879293/

Call Me

+91 8310654643

GitHub

https://github.com/manojaa2003

Python Project

Expense Tracking System

Domain/Function: Software Development

Created a Streamlit–MySQL tracker reducing monthly tracking effort 45% using structured workflows shaped via user interviews. Improved data reliability and displayed pandas insights increasing clarity 40% through refined visual layouts.

View Project
on GitHUB

Analytics by Date

Analytics by Month

Project Details

Overview: Developed an expense management application with a clean and interactive UI powered by Streamlit and a robust backend built with FastAPI.

Frontend: Created an easy-to-navigate dashboard using Streamlit, enabling users to input, view, and manage financial data with real-time responsiveness.

Backend: Engineered a RESTful API using FastAPI to handle data CRUD operations, enforce business logic, and provide a scalable foundation for future enhancements.

Testing: Implemented comprehensive testing frameworks for both frontend and backend, ensuring reliability and ease of maintenance.

Deployment Workflow:
->Install dependencies via pip install -r requirements.txt
->Launch the service with uvicorn server.server:app --reload
->Start the UI with streamlit run frontend/app.py

Key Strengths: Modular architecture (separating UI and API), emphasis on testing, and clean documentation for smooth onboarding.

Credit Card Launch

Domain/Function: Statistical Data Analysis

Performed end-to-end data cleaning, feature engineering, and A/B testing to compare a new credit card variant with the existing one. Used statistical methods to quantify uplift and support data-driven launch strategies.

View Project
on GitHUB

More than 50% of customer base are in in age group of 26 - 48 adn ~26% are of age group 18 - 25

People with age group of 18 -25 accounts to ~26% of customer base in the data,Avg annual income of this group is less than 50k

Project Details

This project focuses on evaluating the effectiveness of a newly introduced credit card variant using real customer transaction data. The workflow includes extensive data preprocessing—handling missing values, removing outliers, validating data quality—and creating analytical features to measure customer spending patterns.
An A/B testing framework was applied by segmenting customers into control and test groups, followed by descriptive statistics and a formal Z-test to measure whether the new credit card led to a statistically significant uplift in spending behavior. Insightful visualizations and comparative metrics highlight how the new variant performs against the existing product.
The project concludes with actionable business recommendations, helping stakeholders assess whether the new credit card variant is ready for full-scale rollout and providing a data-backed foundation for launch strategy decisions.

Machine Learning Project

Health Insurance Premium Prediction

Domain/Function: Machine Learning

This project builds a machine learning model to predict health insurance premiums using demographic, lifestyle, and medical factors. It includes data cleaning, feature engineering, model comparison, and a full end-to-end predictive pipeline.

View Project
Engagement

View Project
on GitHUB

View Project
on

Entering the applicants details

Predicted cost as a output

Project Details

The Health Insurance Premium Prediction project aims to build a complete machine learning system that accurately estimates insurance premiums using demographic, lifestyle, and medical attributes. The process begins with extensive data cleaning, including handling missing values, outliers, and data inconsistencies to ensure a reliable dataset. Detailed EDA is performed to understand relationships between age, BMI, smoking status, region, and premium costs.
Advanced feature engineering techniques such as encoding, scaling, and interaction-term creation are applied to improve model performance. Multiple regression models—Linear Regression, Random Forest, Gradient Boosting, and XGBoost—are trained and evaluated using RMSE, MAE, and R². Hyperparameter tuning further enhances accuracy.
The final output is an optimized ML pipeline capable of generating fast, reliable premium predictions, providing value for insurance companies in pricing strategy and risk assessment.

Machine Learning Project

Credit Risk Prediction

Domain/Function: Machine Learning

A machine learning–based credit risk modeling system that predicts the likelihood of loan default using customer, loan, and bureau data. Includes data preprocessing, feature engineering, model comparison, hyperparameter tuning, and a Streamlit web app for real-time predictions.

View Project
on GitHUB

View Project
on

Entering applicants details

Loan Default Probability as output

Project Details

This project focuses on building an end-to-end Credit Risk Modeling system to predict the probability of loan default using customer, loan, and bureau data. It includes extensive data preprocessing, handling missing values, outlier treatment, encoding, and feature scaling. Advanced feature engineering techniques were applied to create meaningful risk indicators.

Multiple machine learning models—Logistic Regression, Random Forest, XGBoost, and LightGBM—were trained and compared using metrics such as ROC-AUC, F1-score, and precision-recall. Automated hyperparameter tuning was performed using Optuna to maximize model performance.

The best-performing model was deployed through an interactive Streamlit web application, enabling users to input customer details and instantly receive a risk prediction score. The system supports real-world credit screening by improving decision-making, reducing default risk, and enhancing financial transparency.