FIne tunned several hugging face transformer model. Build the citation tool from scratch.Experienced in building datasets, training and finetuning Transformer-based NLP models (e.g., fastText, BERT, T5), and optimizing models for efficiency. Worked on NLP task like Translation, Text summerizer, and other LLM models. A web scraping solution developed by me ended up getting reused for data collecction.Scraped more than 1cr urls within 10 days which include journals, articles and research papers. Also scraped wikipedia

Managed regional flat performance, optimizing P&L and inventory allocation. Skilled in analyzing large datasets and developing predictive models like recommendation systems (Build from scratch and deployed), as well as web scraping and chatbot developmentDeveloped pricing strategies and generated accurate forecast weekly, monthly, and quarterly. Proactively identified peak and shoulder months, leveraging past performance and industry trends. Conducted in-depth analysis on user, supply, and demand profiles to improve revenue trends. Utilized Python and SQL for data mining and analysis, identifying opportunities for productivity enhancement.

Toxic Comments Classification:Utilized advanced NLP models and ensembletechniques to effectively classify toxic comments.Analyzed and processed a dataset of toxiccomments, employing techniques such as WordEmbeddings, TF-IDF, and LSTM.Optimized models for practical deployment in real-world scenarios through Streamlit applications