Data Scientist
Current• Expertise in Machine Learning and Computer Vision, specializing in key point detection, object detection, classification, and segmentation using Custom Models.• Natural Language Processing, utilizing transformer-based architectures for text generation and Named Entity Recognition (NER).• Document Analysis, including Optical Character Recognition (OCR) and key information extraction, with experience in implementing state-of-the-art models and also reading documents using models without OCR.• Experience in Time Series Predictions and Tabular Data Classic Machine Learning.• Ability to integrate Machine Learning models into backend servers using FastAPI and Flask and in integrating Machine Learning models into mobile applications using ONNX Runtime and TFLite.• Experienced with AWS, PySpark, SynapseML, ONNX, and various data processing libraries for building efficient data pipelines.• Developed interactive Gradio interfaces for machine learning model deployment and created backend/frontend solutions in multiple programming languages.• Created tools for efficient data labeling, enhancing the accuracy and speed of the annotation process.• I have implemented and trained various types of models, including encoders/decoders for transformer like models and convolutional networks from scratch. I have also developed new approaches to solve specific problems.