• Working on LLM fine-tuning and RAG/Agentic-RAG based LLM systems.• Worked on e-commerce fraud detection using Graph Neural Networks(GNNs).

• Developed a deep learning API for tracking color changesin Welley's urine dips(18 color markers). Tools: YOLOv11, Python-Colormath, OpenCV, PIL, Scikit-image and FastAPI.• Worked on DNA(50 kilobase bins) data analysis andpredictive modeling. Tools: Scikit-learn, numpy, pandas.• Developed a deep learning API for tracking color changes in Uriwell urine dips(12 color markers). Tools: YOLOv8, Python-Colormath, OpenCV, PIL, Scikit-image and FastAPI.

• Developed an intelligent chatbot and its docker image forbooking of slots for doctor consultation, lab tests andfetching the reports leveraging Google's BERT LLM in thebackend. Tools used: RASA, Docker• Improved pdf diagnostic report processing time throughan OCR-based information extraction pipeline. Tools:pypdf, pymongo.• Designed and deployed a scalable RAG-based chatbotusing LangChain and Pinecone.• Worked on APIs development and testing using Flask and FastAPI. • Worked on data analysis and insights derivation on differentdatasets from company's relational database. Tools:NumPy, Pandas, SQL, AWS and PySpark. • Researched for cutting edge AI applications in generative AI, Large Language Models and Computer Vision Algorithms. Tools used: PyTorch, TorchVision, TensorFlow, LangChain, Assemblyai, Elevenlabs, Llama, BERT, GPT, HuggingFace, SpaCy etc.

• Worked on the Probyto AI flatform Artificial IntelligenceManagement System(AIMS). Tools: Docker, Apache AirFlow.• Worked on predictive and descrptive analytics using tools like NumPy, Pandas, Seaborn.