A dedicated and detail-oriented professional with a strong passion for creative problem-solving and delivering high-quality results. Possessing solid expertise in natural language processing (NLP), machine learning, and hands-on experience with generative AI systems, I am skilled in developing and deploying cutting-edge AI solutions to meet business objectives. having knowledge of industry tools and techniques, coupled with a deep understanding of AI platforms such as ChatGPT and Prompt Engineering
Intelligent Document Retrieval and Summarization System (IDRSS), Developed end-to-end ML solution for legal document retrieval and summarization. Implemented a custom RAG pipeline with BERT and Faiss, reducing retrieval time by 50% and improving relevance by 35%. Fine-tuned GPT-based models for context-specific summarization and question-answering, deployed via FastAPI. Built scalable infrastructure with Docker and Kubernetes, leveraging AWS Lambda and S3., Python, TensorFlow, PyTorch, BERT, Langchain, Docker, Kubernetes, Fine-tune, FastAPI, MLflow, Grafana, AWS S3, Faiss
Turn-Taking and Backchannel Prediction Project, Implemented this research paper https://arxiv.org/pdf/2401.14717. Developed a labeling framework for key conversational elements: 'Backchannel,' 'Turn-taking,' and 'Continuing Speech.' Preprocessed data by normalizing transcripts, managing overlaps, simulating speaker roles, and balancing datasets. Implemented GPT-based models to integrate acoustic and linguistic cues for improved predictions., ASR, Python, Hugging Face, PyTorch, NLP, Machine Learning