Entry-level Data Engineer dedicated to designing, optimizing, and managing robust data pipelines. Strong foundation in P

github

Subham Mohapatra  • Peerlist

Built a supervised ML model to predict diseases using patient medical records, achieving 72% accuracy. • Used Pandas, NumPy, and Scikit-learn to preprocess structured healthcare data. • Applied classification algorithms like RandomForest and Logistic Regression. • Visualized model outcomes and risk patterns using Power BI and Matplotlib. • Integrated with a MySQL backend and presented predictions in dashboard format.

Disease Prediction Model

Created a sentiment analysis system to classify hospital patient feedback as Positive, Negative, or Neutral. • Used NLP tools including NLTK, TF-IDF Vectorization, and VADER Sentiment Analyzer. • Balanced training data with SMOTE and fine-tuned model using RandomForest Regressor. • Integrated model outputs with a MySQL database for real-time feedback analysis. • Results presented in an analytics dashboard with Power BI visualizations.

Sentiment Analysis & Feedback Classifier

Contributed to a real-time energy monitoring system to predict power usage and detect anomalies (e.g., energy theft) in public infrastructure. 🔹 Forecasting Module: • Built ARIMA-based models using Python and statsmodels to predict power consumption trends. • Processed and analyzed real-time data from IoT devices (~3–4 lakh records) using Pandas and MySQL. • Worked with the backend team to integrate forecast results into APIs. 🔹 Anomaly Detection + Frontend Module: • Developed anomaly detection logic using statistical thresholds and energy behavior rules (e.g., sudden spike in current_ph1 without matching load). • Created a responsive frontend dashboard in React.js + Tailwind CSS to visualize power metrics, forecasts, and alerts. • Used Chart.js for graphs, Axios for API calls, and collaborated on UI testing for real-time data monitoring.