Siddharth Kalla

Working in the Procurement and Research team where building possible solutions to optimize the procurement work and taking better trading decisions. Worked on various projects such as building Bi-directional models for predicting the trading signals, build optimized clustering analysis for the different voluntary carbon credit projects and lately worked on building a multi-functional dashboard for all research and procurement purposes for our internal team and deployed it onto GCP.

Worked in the risk management sector for the trading team, developed logics for VaR calculations with performing all risk related analysis in python. Also worked on python automation projects such as auto-mailer and auto-data fetcher. Also assisted in modifying and handling a Data-Science training program “YoDaS” for the company, for young data-scientist aspiring to learn and change their career path. Here I mentored a batch of 15 students and focused on delivering data science concepts and making them understand load forecasting and help in building forecasting models.

Worked on Building different GAN architecture for MNIST image generation. Carried the same architecture and testing it on camoGANs data used for camouflaging military purpose.

Worked on Physiological Data for Classification of raw sensor data for different stress types using deep learning techniques such as MLP, CNN and RNN, LSTM. Understood the working of residual networks and designed ResNet50 from scratch.

Worked in a team dedicated to research about Optical Character Recognizers and using them to extract useful information from legal documents, also worked on deploying the same on mobile application. Also carried the study on understanding Convolution and building CNN layers from scratch which gave a detailed understating about working of neural networks.