->Attention-Based Deep Neural Networks for Low-Resource Indian Regional Language Analysis  Fine-tuning BERT and IndicBERT transformers for Kannada language parts of speech tagging Technology stack: Pytorch, Torchtext library, Python, Deep learning, Tensorflow  In cross validation, it achieved 94% accuracy.-> Brain tumor detection and segmentation from MRI images using deep learning and machine learningtechniques Annotation has been performed on the dataset containing 250 images. Furthermore, by using data augmentationtechniques,thedataset was increased to 2000 images.The model was trained for object detection tasks with techniques like CNN, XgboostonthePARAM Utkarsh supercomputer (Nvidia Tesla v100). For object segmentation implemented using Yolov5, Yolov7. ->Time series forecasting Bengaluru city’s Temperature, Humidity, CO2 level, PM 2.5, and PM 10 particles in the air for the next 24hourshavebeenforecasted.  Implementing cutting-edge time series algorithms such as DeepAR on the MXnet framework, SARIMAX, and Prophet usingAutoTS. The accuracy of 96% was achieved using Prophet.