Ishwar Sawale
NATURAL LANGUAGE PROCESSING & COMPUTER VISION · DEEP LEARNING & MACHINE LEARNING ENGINEER ü +91-82-374*-**** Ó ************@*****.*** ! ishwarsawale.github.io ishwarsawale ƅ courses ishwarsawale Summary
I am a Software Engineer working on Machine Learning & Computer Vision algorithms for 4.2+ Years. I am currently working with Mindstix Software Labs as Solutions Architect Data Science. I am an Engineer focus on data-driven solutions for Deep Learning, Computer Vision, and ChatBot based systems. I have implemented the number of projects from research till deployment in production. Before Mindstix, I was working with Coriolis Technologies Pvt Ltd for two years as a Machine Learning Engineer, developing Machine Learning solutions for the clients. AsSoftwareEngineering is improving steadily, to keep up with the industry, I have also completed 150+ courses related to Machine Learning, Software Engineering, etc. from Udemy, Coursera, Datacamp, & LinkedIn Learning. Professional Experience
Data Science Solution Architect - Mindstix Software Labs VMware Inc GENIE SPEAKER RECOGNITION ù PYTHON, TENSORFLOW, CNN, RESNET, ANN INDEXING, VLAD Dec 2018 - Present
• This is a text-independent speaker recognition systembuiltusingdeepneuralnetwork. It is based on ResNet architecture with NetVLAD
& GhostVLAD layers. This is a three-stage system consists mainly of Development, Enrollment & Verification stage.
• In the Development stage, I have created new modified thin ResNet architecture with 34 layers & 3 million parameters. It is trained end-to-end for speaker recognition on 10GB of speaker data for 8K Classes using Adam optimizer with additive margin softmax
(AM-Softmax) loss.
• In the Enrollment stage, createdaspeaker-specificmodelusingembeddingsextractedfrompreviouslytrainedmodel.In the Verification stage, used high dimensional Approximate Nearest Neighbour Indexing to predict nearest match for test input. Machine Learning Engineer - Mindstix Software Labs VMware Inc. GENIE DIGITAL ASSISTANT ù PYTHON, TENSORFLOW, KERAS, LSTM, RASA Feb 2018 - Aug 2019
• This application is a digital chatbot assistant helping customers to get their issues resolved. Built the Natural Language Understanding framework for entity recognition, intent classification & dialogue management using LSTM architecture along with other libraries like Spacy, Rasa, etc.
• Created Frequently Asked Question service using Universal Sentence Encoder Computer Vision Engineer - Mindstix Software Labs Estée Lauder Companies Inc. FOUNDATION RECOMMENDATION ENGINE ù PYTHON, OPENCV, COMPUTER VISION, DEEP LEARNING, TENSORFLOW Feb 2018 - Dec 2019
• Created a new Fully Convolutional Network using AlexNet architecture with a domain-specific loss function.
• Using developed algorithm, color corrected image is generated basis which skin tone of a person is calculated using statistical methods.
• Built Product recommendations based on the user’s preferences & extracted skin tone. Machine Learning Engineer - Coriolis Technologies
LICENSE PLATE RECOGNITION SYSTEM ù CNN, OPENCV, PYTHON, OBJECT DETECTION, CLASSIFICATION, YOLO March 2016 - Feb 2018
• An Automatic License Plate Recognition System using Deep Learning algorithms for Indian vehicles.
• I have trained 5 different YOLO CNN Networks for vehicle detection, prediction of the bounding box for License Plate, character segmen- tation, alphabet recognition & digit recognition.
Skills
Languages & Frameworks Tensorflow, Keras, Pytorch, Python, Django, Flask, Pandas, Numpy, Matplotlib, GO, Ruby, Rails Machine Learning Regression Algorithms, Decision Trees, SVM, Navie Bayes, kNN, K-Means, Random Forest etc. Deep Learning CNN, ResNet, DenseNet, Inception, MobileNet, RNN, LSTM, Chatbot Computer Vision Classification, Object Detection, Segmentation, Color Correction, Image Processing MARCH 26, 2020 ISHWAR SAWALE · RÉSUMÉ 1
Education
Diploma in Big Data Analytics Pune, India
CDAC-ACTS PUNE(70%) Aug 2015 - Feb. 2016
Bachelor of Engineering Pune, India
SINHGAD ACADEMY OF ENGINEERING, PUNE (69.33%) Aug. 2011 - May. 2014 Courses List
PyImageSearch
• PyImageSearch Gurus Graduate by PyImageSearch
Coursera
• Mathematics for Machine Learning, a 3-course specialization by Imperial College London
• TensorFlow in Practice Specialization by deeplearning.ai
• Structuring Machine Learning Projects by deeplearning.ai
• Neural Networks and Deep Learning by deeplearning.ai
• Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization by deeplearning.ai
• Recommender Systems: Evaluation and Metrics
• Nearest Neighbor Collaborative Filtering
• Matrix Factorization and Advanced Techniques
• Introduction to Recommender Systems: Non-Personalized and Content-Based
• Fundamentals of Digital Image and Video Processing by Northwestern University
• PCA by Imperial College London
• Multivariate Calculus by Imperial College London
• Linear Algebra by Imperial College London
• How Google does Machine Learning by Google Cloud Datacamp
• Machine Learning with Python Track
• Data Scientist with Python Track
• Image Processing with Python Track
• Data Manipulation with Python Track
• Data Analyst with Python Track
• Importing & Cleaning Data with Python Track
• Data Scientist with R Track
• Data Analyst with R Track
• Importing & Cleaning Data with R Track
• Machine Learning with R Track
MARCH 26, 2020 ISHWAR SAWALE · RÉSUMÉ 2