I’m particularly interested in applied deep learning for mental health, focusing on Natural Language Processing (NLP) and Speech Emotion Recognition (SER). My work includes fine-tuning transformer models like BERT, RoBERTa, and DeBERTa on emotion-labeled datasets such as GoEmotions and audio-based models for depression detection. I’m also exploring multimodal fusion approaches that combine text and audio to build intelligent, real-time companions for mental wellness support.