Join the EduLinkUp Team

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.

House Price Predictor

by Anisha Shaw

This project focuses on predicting house prices using basic machine learning techniques in Python. It is designed to be easy to understand, especially for beginners learning data science and ML. Built a house price prediction system using Python and machine learning. -Performed Exploratory Data Analysis (EDA) to understand feature relationships. -Handled missing values and applied feature scaling & encoding. -Trained a Linear Regression model using scikit-learn. -Evaluated the model using RMSE, MAE, and R² score. -Visualized results with actual vs predicted price plots. -Saved the trained model for future use or deployment.

Security Report Generator

by Eishit Balyan

I built a security vulnerability dashboard that takes Nessus scanner CSV files and actually makes them useful. when you run a vulnerability scan you get a massive CSV dumped on you and figuring out what to fix first is genuinely painful. So the app ingests that file, parses out all the assets and vulnerabilities, stores everything in a proper database, and then automatically calculates a risk score for every asset. The formula multiplies each vulnerability's CVSS score by a severity weight and rolls it up by asset criticality — so you immediately know which machines are most exposed and why. On top of that it hits the National Vulnerability Database API in the background for every CVE it finds. So within about thirty seconds of uploading a scan you've got CWE classifications, official CVSS scores, patch availability, and full descriptions pulled from NVD automatically — no manual lookups. The frontend is a React dashboard. There's a risk ranking table where you can click any asset, see every CVE on it with direct NVD links, and adjust its criticality score in real time. Charts, severity breakdowns, scan history — all of it updates without page reloads. The feature I'm most proud of is the one-click PDF report generator. Click export and you get a proper multi-page PDF — branded cover page, executive summary written from actual data, severity charts, full findings breakdown, and auto-generated recommendations based on what was found. Something you could actually hand to a manager or auditor. Stack is FastAPI, SQLAlchemy, SQLite on the backend and React with plain CSS on the frontend.

Titanic Survival Predictor

by MALLESWARAPU SRIYA

I built a machine learning model that predicts whether a Titanic passenger would survive based on features like age, gender, passenger class, fare, and family information. The project involved data preprocessing, handling missing values, encoding categorical variables, and creating new features such as family size and passenger titles. I trained and compared multiple models including Logistic Regression, Decision Tree, Random Forest, and SVM, and selected Random Forest as the best-performing model after hyperparameter tuning. Finally, I deployed the model using a Streamlit web application where users can enter passenger details and get real-time survival predictions.

Auto-Scaling Web Service

by Kavinaya B S

Designed and deployed a highly available and scalable web application using Amazon EC2, Application Load Balancer, and Auto Scaling Groups with CPU-based target tracking policies. Implemented dynamic scaling to automatically increase instances during traffic spikes and reduce capacity during low demand, ensuring performance and cost optimization. Integrated CloudWatch monitoring and SNS notifications to track scaling events and system health across multiple Availability Zones.

Diabetes Risk Assessment

by Tanuja Sandip Nalage

This project builds a Diabetes Risk Assessment system using machine learning on the Pima Indians Diabetes dataset. The workflow includes data cleaning of medically invalid zero values, median imputation, feature engineering (BMI categories, age groups, glucose levels), and standardization. Multiple classification models (Logistic Regression and Naive Bayes) are trained and evaluated using accuracy, precision, recall, F1-score, and confusion matrices. The project emphasizes recall as a critical metric for medical screening to minimize false negatives (missing diabetic patients). Health insights are derived from key risk factors such as glucose level, BMI, age, and insulin patterns. A trained model is saved and the complete implementation is documented with medical disclaimers and ethical considerations.