Vishwas
Chhimpa

• Forecast multiple sales and adherence time series using ensemble Machine Learning algorithms considering sparse and short history sales data while handling seasonality and holiday trends.

• Leveraged technologies like CatBoost/XGBoost, LLMs, etc., to create deliverables for client sales strategies.

• Created practical sales force strategies that generated higher sales, detailed multiple products, supported new product launches, and defended against new competitive product launches, directly impacting a positive new profit of $15 million for one business unit.

• Created required automation based on client requirements in data workflows from data ingestion, quality checks, processing, modeling, to final data creation and reports using AWS, Snowflake, Salesforce, and PowerBI.

MVP product, Firebase, GCP, and Vue.js, integrating APIs like Places API, Email API, Google Analytics, and SMS API.

Generated the 24000 simulated data of pressure at nodes from generating leaks in pipe network using Epanet simulator and python. Applied Xgboost algorithm to create a model for classification of leak location and used a regression model for the leak size prediction.

Processed 1200 images of two concrete samples involving denoising, masking, and filtering of images using OpenCV. Published a paper in the 2019 edition of the 7th ICCMS conference at IIT Mandi, selected for publication on Springer – “Online ISBN978-981-15-8315-5“.

Developed an Indian Sign Language hand gesture recognition system using Image processing and Convolutional Neural Networks. The final model classified images with an accuracy of 73.5%.