Utkarsh Shinde

I. Project – Customer Service Chatbot using Large Language Model

• Designed and developed an LLM-based chatbot using Llama 2 to enhance customer service operations.
• The chatbot effectively managed customer queries, provided support, and automated routine tasks, resulting in reduced response times and operational costs.
• Optimized the model through fine-tuning with domain-specific data, also tested and refined the chatbot's performance using various evaluation metrics and user feedback resulting in a 25% increase in user satisfaction scores.
• Successfully integrated the chatbot with the customer service platform, enabling seamless interactions with customers

II. Project – Dynamic Real Estate Price Predicting Tool

• Collected and preprocessed a large dataset of real estate listings, including features such as property type, location, size, and amenities
• Developed and trained multiple machine learning models, including linear regression, decision trees, and neural networks, to predict property prices
• Implemented feature engineering techniques, such as data normalization and feature scaling, to improve model performance
• Conducted exploratory data analysis to identify trends and patterns in the data, and to inform feature selection and model development
• Deployed the model using a web application, allowing users to input property characteristics and receive predicted prices

III. Project – E-commerce Recommendation System

• Collected and preprocessed a large dataset of product description, product price, product category, purchases, and ratings
• Developed and trained multiple machine learning models, including rating based, content-based, collaborative filtering, and hybrid approaches, to generate personalized product recommendations
• Conducted exploratory data analysis to identify trends and patterns in customer behavior, and to inform feature selection and model development
• Integrated the recommendation system with a e-commerce application, that achieved a 25% increase in customer engagement and a 15% increase in sales