Created pipeline for automatically doing anomaly detection of time series data collected across the servers. Coordinated statistical data analysis, design, information flow and performed documentation, and implementation of the Cisco Webex server data. Built library of models and reusable knowledge-base assets to produce consistent and streamlined business intelligence results for doing time series forecasting of Webex data. Used statsmodel, NumPy, pandas, Matplotlib, Sklearn for data analysis and plotting. Link:- https://github.com/kampaitees/Internship-Project

A research paper that proposes a new design strategy for Capsule Network architecture for efficiently dealing with complex images. It outperforms the top-5 performance on benchmark datasets CIFAR-10 and Fashion MNIST with highly competitive performance on the SVHN dataset. The proposed method incorporates wide bottleneck residual modules and the SE attention blocks to address the defined problem. A wide bottleneck residual module facilitates extracting complex features whereas, the SE attention block enables channel-wise attention by suppressing the trivial features.

Designed and developed a web portal for doing criminal image searches from their database using Facial Recognition. Used Django as backend, python as a scripting language and libraries such as Numpy, Pandas and Keras for machine learning and data analysis. Analyzed crime data to give insights about the crimes happening, showing the correlation between the places and particular crimes to reduce them.