Worked as the Google Cloud Consultant to provide solutions for Customers who are using Google Cloud. Delivered training for working professionals in the project, took many initiatives such as Creating Break Fix lab, Mock demo over video call with. Which improved the troubleshooting approach and Project performance.Collaborated with SMEs across various domains to identify training needs and skill gaps,thereby customizing sessions to precisely align with organizational goals and objectives.

Delivered Customised training onsite and online on Google Cloud Certifications such as Cloud DigitalLeader, Associate Cloud Engineer and Professional Cloud Architect.Created customised outlines and content for various Google Cloud Certification,Python, Advanced Python Machine Learning and Deep Learning.

Face Recognition and Drowsiness DetectionBuilt an app that detects the attendance and attentiveness of the online classroom using live video from the webcam and real-time aggregated feedback to the instructors about the class. Employed image preprocessing techniques like image augmentation, Pixel brightness transformations to improve the quality. Deployed quantized model on AWS using Streamlit API that alerts drowsiness of attending students class. High quality image require for training of model. Experimented with multiple libraries such as face_recognition and dlib to extract facial embeddings and landmarks, the final model built with an average accuracy of 88%. https://github.com/BalajiJadhav015/Face-Recognition-and-Drowsiness-Detection.gitTopic Modeling on BBC News Articles.Implemented LDA topic model algorithm on the content of the news articles to discover major themes such as politics, business, sports and technology. Performed text preprocessing in spaCy like Stemming, tokenization and vectorization using TF-IDF vectorizer, named entity recognition. With help of Latent Dirichlet Allocation able to find best model parameters from range of (5,10) from pyLDAvis visualized number of topics. Preformed label encoding on the News data experimented many models like Random Forest Classifier, Decision Tree Classifier, KNN classifier train and test accuracy as 0.96,0.95 with average f1 score of 0.95.https://github.com/BalajiJadhav015/Topic-Modeling-On-News-Article.git