Roshan

PROJECTS Medical OCR system ● Build a Smart OCR System to read a variety of medical forms ( printed and handwritten text ) and mediation prescription ● Developing Drug Drug interaction Model based on Drug catalog open sourced by FDA ● Early Disease Detection in herds by image classification via GoPro images based inside cattle ranches, directly correlated to cattle medication treatments AI Assistant for platform tour and actions ● Designing a chatbot interface to guide users through complicated web services platforms by navigation and interactive command prompts ● Building a chatbot interface based on NLU and NLG systems using a Bi-RNN model running inside a Sanic loop containerized serverlessly. ● Developing mongoDB knowledge base connections to the chatbot server to inherit dynamic Building Complex Imaging systems on TFS ● Building stateless python servers divided between microservices ● Constructure TensorFlowServing for stateless cloud build solutions, allowing complex tensor compilers Reddit Sentiment Analysis Examination conducted by Microsoft for validation of skills and industry presets required ● End-to-End Machine Learning with TensorFlow on GCP (08/2020) coursera.org/verify/7B M64VLTXCGH ● Reliable Google Cloud Infrastructure: Design and Process (07/2020) coursera.org/verify/JXT JTFR4WC2N ● ISRO - Satellite Photogrammetry and its Application (08/2020) 3a1d419578836522c28f a6f3a77b6cf1 ● The project uses Reddit's classifications (subreddits) and quantitative data ● compared, to provide a better collective understanding of large masses ● Using NLTK to target specific demographic and provide real time Sentiment from ● reddit users ● E-Comm Sentiment Analysis and Time Series ● Scrapping real time data from 50 E-Comm sites of products on a daily basis ● (Automation) ● Scrapping reviews and creating category wise sentiment break down (Sentiment ● Analysis) ● Searching for Time-series patterns and anomalies Facial Detection Attendance ● Facial Recognition for attendance calculation in Real Time via Online classes ● (Zoom/Google classroom) ● Processing student/employee faces and classification into selected emotions for ● proper post analysis of group mental and emotional status. ● Using Google Cloud to seamlessly integrate the various processes. ● Provide a data Visualization of the group in an interactive DashBoard via Tableau.