Leveraged Machine Learning and Deep Learning models for battery life prediction and automation.
- Achieved 90% accuracy in Remaining Useful Life (RUL) predictions for lead-acid batteries using Random Forest, XGBoost, LSTM, and CNN.
- Preprocessed raw laboratory data, engineered features, and implemented classical ML and Neural Network architectures.
- Developed CNN-based image classification models and built prototype RAG-powered chatbots using OpenAI and Pinecone.
- Automated complex workflows using n8n for improved research efficiency.
- Tech Stack: Pandas, Numpy, Scikit-learn, TensorFlow, PyTorch.