SAATWIK BISARIA

Proven experience in developing end-to-end data science pipelines and products to optimize business processes and generate value for the organization. Projects include: DG Classifier (NLP-based pipeline mitigating 99% DG shipment risk and saving ~300Cr Rs Daily); Grid ID Classifier (hierarchical classification pipeline reducing shipment misrouting by 5% and saving ~20L Rs Daily); and Storefront Program Pan India Recommendations (data-backed validation for Ecom Storefront Partner initiative).

Developed an end-to-end hierarchical classification pipeline for product classification (Product Classifier). Developed an end-to-end system for recommending ideal facility locations (Facility Location Planning). Created a system dividing the Indian subcontinent into micro-grids and macro-grids (Pan India Grid System).

Built End to End DG Classifier Product for classification of shipments into dangerous goods using NLP and Gradient Boosting Techniques. Performed Data preparation, Text preprocessing, Feature engineering, Train and Test models, Data visualisation and Django API Development and Deployment.

Presented report on working prototype for DexConnect platform and Performed analysis and review of all Dex platforms.

Developed house price predictions using Artificial Neural Networks pipeline and wrote research paper on the project for AI Tech Summit 2019.

Developed an End to End Pipeline to find out the views of different people on the demonetisation by analysing tweets using AFINN dictionary (Sentiment Analysis).