Machine Learning Engineer at Rightance Healthcare Inc

Developed deep learning models for medical image classification and segmentation on a 60K+ clinical image dataset. Improved diagnostic signal detection and reduced label noise by 22% through data augmentation and preprocessing pipelines. Built automated training, validation, and evaluation pipelines tracking model performance using AUC, F1-score, precision and recall, improving inference efficiency by 30% while maintaining clinical-grade accuracy. Deployed containerized model inference services using Docker and REST APIs on Azure, enabling scalable integration with clinical imaging pipelines and real-time model inference.

AI Engineer / Data Science Intern at Kapital Bank

Built a real-time conversational banking assistant using GPT-4 via Azure OpenAI, integrating STT/TTS pipelines to support 100K+ banking users, reducing customer support response times by 20%. Designed WebSocket-based streaming pipelines for real-time STT/TTS processing, enabling low-latency conversational interaction with banking services. Implemented a Retrieval-Augmented Generation (RAG) architecture, retrieving contextual information from internal banking knowledge bases via REST APIs to generate accurate and compliant responses.