Shrinidhi Kumbhar

Master's

Artificial Intelligence and Robotics

Brightleaf Solutions

Data Science Engineer (NLP Research Engineer)

• Collect, clean, process, visualize and analyze data before modeling using python • Develop ML(machine learning, Deep Learning) models to extract information from different document formats such as • PDF, JPEG, and PNG using computer vision, NLP, and OCR with python • Train models using Nvidia GPUs such as Tesla T4, Nvidia GTX 1060, and GPU instances available on AWS Sagemaker, and fine-tune these models to achieve higher accuracy • Work with algorithms such as Faster-R-CNN, YOLO, and SSD for object detection, and use Tesseract OCR tool for text recognition and extraction • Tag raw textual data into Part-Of-Speech and Inside-Outside-Beginning format using python • Develop DL models using architectures such as BERT, RoBERTa, DistilBERT, RNN, LSTM, and GRU for text classification, sentiment analysis, • Develop named entity recognition models using DL and libraries like spacy, nltk and flair, to extract specific textual data from documents • Developed a synonyms recommendation application using GloVE 300d embeddings at the backend with python • To update existing ML, DL models on new data, implemented online/incremental machine learning • Work with TensorRT to optimize trained ML, DL models for run-time performance and handle database with MySQL and MongoDB