Where I've worked

Research Collaborator

• Developing a mathematical framework for handling long-tailed distributions in multimodal 3D perception.
• Emphasizes feature-space robustness and fair representation of rare classes.
• Validated on state-of-the-art autonomous driving benchmarks.

Research Assistant

• Conducted research on the application of Liquid Neural Networks for ECG Abnormality Detection, focusing on improving interpretability and robustness.
• Developed a novel deep learning pipeline to classify ECG signals with high accuracy, leveraging time-series modeling techniques.
• Submitted a research paper titled "Learning Interpretable Cardiac Dynamics with Continuous-Time Neural Models" under review at Annals of Biomedical Engineering.

Data Science Intern

• Designed and developed an NLP-based chatbot for automating customer support, improving response times and satisfaction.
• Fine-tuned a Transformer-based model (BERT) achieving 90% accuracy in intent recognition.
• Deployed the chatbot to production on AWS with scalability and integration into existing support infrastructure.

Selected Participant

• Participated in an intensive program focused on Natural Language Processing.
• Developed a Text-to-Speech system for multiple low-resource Indic languages, enhancing accessibility.
• Presented the project to peers and faculty, receiving positive feedback for innovation and implementation.