VIKAS KUMBHARKAR

Design, Develop and Deploy entire Data science life cycle using NLP and data processing pipelines. Ensuring problem definition, data exploration, modeling, analysis, and deployment to production servers. Industrializing data feeds and creating pipelines into existing data systems. Facilitating connectivity of data and applied technologies between internal and external data sources. Ensuring NLP models can be deployed and monitored in a quick, reliable and robust manner. Key achievements include building keyword extractors using sentence transformer and scispacy, and developing data visualization using streamlit from Apache SOLR.

Developed OCR and NLP pipelines for text extraction from various document types. Utilized Google Cloud services for extracting text from PDFs, images, and word documents. Designed, tested, and deployed NLP models for NER and Text Classification using CNN and LSTM models. Automated rule-based systems by interacting with operations teams.

Researched and simulated various Computer Vision techniques, including developing and optimizing preprocessing of DICOM Images. Trained and evaluated deep learning models for image classification tasks.