I am Satya Sri Virinchi Junuthula, a dedicated professional specializing in machine learning and data structures and algorithms. Currently pursuing a Master's degree in Information Technology from Arizona State University, my passion lies in leveraging advanced algorithms to drive innovation in the field.
Top2Vec: I utilized top2vec, a technique that combines word embeddings and HDBSCAN clustering, to preprocess COVID-related data obtained from PubMed. My objective was to determine the optimal number of topics through semantic clustering, facilitating thorough topic identification and enabling detailed analysis
Clinical Trial Chatbot: Built an end to end clinical trial chatbot, which books an appointment for cancer patients. Implemented using NVIDIA RIVA Transfer Learning Toolkit to solve the Intent-Slot classification problem, using custom phrases for training. It has multiple intents, entities, and responses and also has a knowledge base.
Pfizer project: Used Pubmed database and NVIDIA NeMo megatron pretrained models to solve Named Entity Disambiguation problem to find the disambiguation for commonly confused abbreviations in medical prescriptions.
Fox Sports: Automatic Speech Recognition(ASR) model evaluation is implemented on the generated text and the original text collected from Fox Sports web pages.
Training: Worked on computer vision problems and used various ML algorithms. Completed Nvidia - Deep Learning Institute’s Accelerating end-to-end data science workflows. Completed Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning courses.
Whatsapp Chatbot builder: I leveraged Ruby on Rails to create a powerful web application that enables customers to build and deploy personalized chatbots for WhatsApp, catering to their business needs. The application encompasses an integrated testing environment, seamless integration with third-party APIs, support for multiple bots, multi-language functionality, and real-time chat transfer to agent groups.
Chatbot - Google Assistant: Developed a robust chatbot using Google Dialogflow to address queries based on specific requirements set by a university client. Implemented an intuitive user interface with options for text, voice commands, and buttons. Launched the chatbot into production through Google Assistant, extending accessibility to 215 countries via various platforms including mobile devices, Google Home, Android TV, Wear OS, and more.
