First technical hire for the startup.Managed Custom Website Chatbot:• Implemented fully website based chatbot using Next.js and LangChain with OpenAI GPT models for natural language processing.• Integrated Pinecone for vector database functionality, enabling semantic search and context-aware conversations.• Optimized the chatbot for performance, delivering high responsiveness and accurate user query handling.Created AI-Powered CMS for High-End Club:• Architected and developed a robust CMS platform using Next.js to manage club operations, member activities, and events.

Designed and implemented machine learning models for emotion classification, leveraging state-of-the-art techniques in both traditional and deep learning. Built multiple classification models, including Support Vector Machines (SVM), Random Forest, and advanced Deep Neural Networks (DNNs) for accurate emotion recognition from visual data. The backend was developed using FastAPI, ensuring high-speed inference and seamless integration with the machine learning pipelines.Specialized in utilizing Transfer Learning to enhance model performance, deploying pre-trained architectures such as ResNet, and InceptionV3, fine-tuning them on domain-specific emotion datasets to achieve superior results. Applied advanced techniques like data augmentation, dropout, and batch normalization to optimize model generalization and prevent overfitting. Developed the frontend using React, implementing dynamic visualizations of emotion predictions and user interaction analytics. Authoring a research paper on the project, detailing the novel approaches in model design, transfer learning adaptation, and the application of these techniques in real-world emotion recognition systems for publication in peer-reviewed journals.