Projects
This cutting-edge Chatbot revolutionizes educational inquiries! Leveraging the power of GenAI and RAG concepts, it optimizes query comprehension and response accuracy, achieving a remarkable 40% improvement in user engagement. Through strategic fine-tuning and prompt engineering, it enhances Large Language Models to ensure every interaction is precise and contextually relevant. But that's not all! Orchestrating LangChain integration, it seamlessly links OpenAI's LLM (ChatGPT-4) with Astra DB, creating a serverless framework that streamlines semantic search processes. With this project, it's redefining the way users interact and learn in the digital age.
In today's digital landscape, social media platforms like Twitter have become vital arenas for discussing diverse topics, including economics. This project explores economic conversations on Twitter, focusing on the #Econtwitter sub-network. By analyzing interactions, detecting communities, and highlighting popular themes, it aims to offer insights into economic discourse. Through natural language processing and cluster analysis, dominant topics are identified, enriching understanding. Network analysis visualizes user interactions, while a Support Vector Machine classifier predicts tweet clusters with 60% accuracy. The findings have the potential to inform public discourse, shape policy discussions, and enhance societal understanding of economic matters.
In the era of data-driven decision-making, market and corporate analysis has evolved significantly, with researchers delving into sample data from large firms. This analysis hinges on factors such as monetary performance, business nature, and market risks. In line with this, the focus will be on the Fortune500 dataset from 2017, ranking the top corporations by revenue for their respective financial years. This report aims to conduct feature selection and compare various regression models—Multivariable Linear Regression, Polynomial, Ridge, and LASSO. Additionally, it will predict profit values and evaluate model performances and accuracy against original profit values. This dataset offers insights into exploratory analysis of companies, addressing questions like the role of ranking in profitability, the impact of employee strength on company revenue, and whether profit is the sole determinant of stock price estimation.
