Data Analyst Machine Learning

NAMRATA NYAMAGOUDAR

Contact: 949-***-**** Irvine, CA 92612 Email: ********@***.*** www.linkedin.com/in/namrata-nyamagoudar-07a694176 https://github.com/NamrataNyam

EDUCATION

University of California, Irvine Irvine, CA

Masters of Data Science GPA:3.9 Sept 2023 – Dec 2024 Related Coursework: Machine Learning, Statistical Methods for Data Analysis, Introduction to Probability and Statistics I & II, Database management and system, Artificial Intelligence, Machine Learning with Generative Models, Deep Learning, Neural Networks, Transformers, Generative AI, Large Vision Foundation models (Vision Transformer, ResNet), Reinforcement Learning KLE Technological University Hubballi, India

BE Computer Science and Engineering CGPA:3.56 Aug 2017 – Aug 2021 Related Coursework: Computer Vision, Machine Learning, Data Mining and Analysis, Applied Statistics, Data Structures & Algorithms, Database management systems, Big Data Analytics, Social Network Analysis, Linear Algebra, Natural Language Processing TECHNICAL SKILLS

Python

SQL

R

C

C++

HTML

CSS

JAVA

C#.NET

Javascript

Azure

Azure Devops

Kubernetes

CUDA

Microsoft excel

OpenCV

Git

Tensorflow

Pytorch

Keras

Tableau

Provenir

MSSQL

MySQL

PoweBI

Numpy

Scipy

Pandas, Scikit-

Learn

Linux

WORK EXPERIENCE

Deloitte Consulting USI Application Modernization and Innovation, Core Business Operations Cloud Developer Bengaluru, India Sept 2021 – July 2023

Contributed to the configuration and optimization of a Commercial Off-The-Shelf (COTS) decision engine to oversee Provenir applications.

Worked on migration of Provenir Desktop(local) and MyProvenir web applications to the Azure Cloud platform aimed at improving scalability, security, and collaboration, utilizing Azure DevOps function apps. Worked on setting up Azure service bus to connect Provenir applications to the decision engine.

Played a key role in enhancing feature deployment efficiency and security by designing specialized Azure DevOps pipelines tailored for rapid and secure deployment of advanced features in higher environments. Contributed to the javascript based application modernization and maintaining Provenir web application hosted on Azure app service.

Azure service bus Az DevOps pipelines Provenir Azure app service Azure function apps Javascript Cloud migration Spookfish Innovations Pvt Ltd.

SDE Intern Bengaluru, India March 2021 – July 2021

Built an object detection and tracking model to detect anomalies and foreign objects for tablet packaging machine, in a self-directed environment.

Developed YOLOV4-DeepSORT framework for multi-object detection and tracking, employing Cosine metric learning to track identical tablets/capsules and foreign objects.

Achieved an inference rate of 0.28 seconds per frame, facilitating real-time detection at 20 frames per second and significantly improving the tablet packaging process upon system integration.

YoloV4 Deep Learning Tensorflow Object detection/tracking Image processing Data filtering Cosine metric learning Computer vision

Research Intern

PROJECTS

Computer Vision and Graphics - 2D Inpainting of Heritage Site Images towards 3D reconstruction Deep Learning, GAN, RCNN, Image segmentation, Image processing, CNN, Data collection/Filtering, Computer Vision, Semantic Image Segmentation, Data Synthesis

Created a deep learning framework that removes occlusions from heritage site images and generates realistic images by inpainting with neighboring pixels.

Utilized Mask-RCNN for instance segmentation, Resnet-101 for feature extraction and a GAN-based approach for inpainting, resulting in a peak signal- to-noise ratio of 34.63.

IITD-AIMS, Computer Vision and Graphics – Project Title: Vision based Techniques to evaluate the Effectualness of Micro Suturing performed by trainee neurosurgeons Computer Vision, HOG-SVM, Machine Learning, Image processing, Research project

Developed a computer vision algorithm to assess the effectiveness of micro-suturing procedures by analyzing input images, providing a score based on criteria including inter-suture distance, sutures per box, suture slack, suture symmetry, and suture angulation.

Used HOG-SVM classifier to train the sample images captured in different magnifications. Achieved Mean Average Precision(mAP) of 67.34% for an IoU of 0.3.

Prediction of daily stock movements on the US market Data Visualization, Data cleaning, Data Mining, Neural Network, Data Analysis, Data Featurization

Developed a deep learning model using KDD process including neural network model, to predict price changes for 700 stocks over a 700-day interval. Attained a 34th rank among 120 competing teams.

Real time Traffic sign board detection Computer Vision, Machine Learning, Transfer learning, ADAS, CNN, Image processing

Utilized transfer learning for real-time traffic sign detection and qualified for the first two rounds of the Smart India Hackathon while creating a driver warning system towards advanced driving assistance system. Built framework using deep CNN - tiny YOLO for signboard localization and MobileNet for classification of signboards.

CERTIFICATIONS : AZ-204: Developing Solutions for Microsoft Azure, Microsoft Azure Developer Associate, AWS Fundamentals: Going Cloud Native, Deep Learning in Computer Vision, Advanced Deployment Scenarios with TensorFlow, Improving Deep Neural Networks: Hypermeter Tuning, Regularization and Optimization

Contact this candidate