Worked on customized research paper implementations to build an efficient data ingestion and tracking component, coding was done in Python.Fine-tuned tree models such as XGBoost with Optuna and deployed on GCP (Google Cloud Platform) for recommendation systems. Built LLM-based applications with LangChain utilizing LLama2 and Gemini models, mainly for translation and agentic products. Built computer vision models for image segmentation and object detection, mainly used for surveillance. Utilized YOLO models, specifically.Written technical articles about machine learning and deep learning along with web scraping tools. Published an article about a fully functioning active learning pipeline for text classification with deep learning on HuggingFace's blog: https://huggingface.co/blog/opinion-classification-with-kiliDeveloped web scraping tools as stand-alone and web applications (REST APIs) using FastAPI and Flask.Built training pipelines for natural language processing models using HuggingFace and Ray.Developed interactive dashboards with Power BI and Plotly Dash.