Trained and Built a Deep Neural Network model using 6976 mridangam tracks from the MIRDATA library for stroke identification. Read more than 10 research papers from websites like IEEE, ResearchScholar, Arxiv, etc., in 2 months and assigned team roles to a team of 4. Used 4+ Python libraries in the project pipeline, such as Tensorflow, sci-kit-learn, librosa, essential and MIRDATA etc, for data loading, training, and classification.

Developed an NLP model to categorize resumes based on keywords into 6+ job categories to benefit hiring for recruiters, trained on 15+ resumes. Utilized 3+ Python Libraries for processing, extraction, and classification, such as Spacy, Pyresparser, sci-kit-learn etc. Actively participated in the AMEX Credit Fault prediction competition in Kaggle and achieved a 0.65 score with the best score of the competition being 0.80.