Vinayak Rai
Designer + AI Enthusiast, Data Scientist, Learner, Programmer, and Designer
Hi there
I'm Vinayak Rai, a versatile full stack web developer and designer pursuing a degree in Data Science and AI at Indian Institute of Information Technology Dharwad. With expertise in the MERN stack and a keen eye for design, I craft seamless web experiences. My passion extends to AI, specializing in machine learning, computer vision, and NLP.
I've completed four internships, published two research papers in IEEE TALE 2024 and T4E 2024, and was selected for the Amazon Machine Learning Summer School '24. My research spans speaker diarization and metacognitive regulation in learning environments. As Tech Team Lead for our DS&AI Society, I've led projects from online platforms for the specially abled to AI-powered attention tracking systems and energy-efficient 3D CNN architectures for lung cancer nodule analysis.
Proficient in Python, TensorFlow, PyTorch, and big data technologies, I'm adept at developing AI solutions and optimizing deep learning models for real-world applications with a focus on energy efficiency and practical impact. My skills extend to full stack web development, where I create dynamic, seamless experiences. Always eager to learn, I thrive in fast-paced environments, pushing the boundaries of AI, data science, and web development.
Send me a message
- Research Associate - Dept of Educational TechnologyMay 2023 - September 2024• Developed AffectBots: A web-based tutoring system using multimodal analysis (audio, video, text) to analyze 100+ hours of student engagement data with 88% accuracy in emotion recognition.• Implemented speaker diarization using PyAnnote and Whisper Speech Recognition model for Indo-English collaborative learning environments, analyzing 140+ hours of student interactions.• Outperformed existing speaker diarization models by 65% on a 200-hour benchmark dataset of classroom recordings.• Research on improved diarization pipeline accepted at IEEE TALE 2024, focusing on its impact on understanding collaborative learning dynamics.• Applied Conditional Random Fields (CRFs) and machine learning to identify triggers of Socially Shared Metacognitive Regulation (SSMR) episodes in collaborative learning.• Developed an automated analysis system for large-scale collaborative learning datasets, with findings accepted for publication at EdTech Society T4E 2024.
Emotional Monitoring of Crypto Traders
Developed a cryptocurrency trading platform that monitors emotional well-being and provides personalized interventions to enhance user decision-making.
Online Schooling for the Specially Abled
Developed an adaptive online learning platform for specially-abled students using computer vision, winning first place in a hackathon for building it in 48 hours.
Centio.Ai Conversational AI with Research and Task Automation
This is AI-driven platform that enables versatile chat interactions, task automation, and in-depth research capabilities. With features like Regular Chat, Assistant tasks, and document management, it streamlines workflows for enhanced productivity.
Alumni Connect
Designed a Flutter app for CGC IIIT Dharwad that fosters alumni engagement, featuring social media-style image sharing, job postings, and job application capabilities.