Sign in

Data Scientist

Seattle, WA
April 25, 2021

Contact this candidate


RUSHABH SHAH Rush**** 682-***-**** San Jose, CA rush1805 Rush1805


Tesla Inc. · Data Science Intern - Supply Chain 08/2020 - Current Developed ARIMA and Exponential smoothing models in R to accurately make forecast recommendations of car parts resulting in improved cost efficiency and an increased customer order fill rate by 17% Deployed an end-end fully automated machine learning tool using Python to alter inventory stocking levels based on predicted consumption and feedback from Part Advisors Built a Decision Tree to prioritize material planning and inventory allocation to support global Model S product launch across 392 Service centers; applied feature-engineering and achieved accuracy rate of 87% Determined optimal order quantity by implementing sampling and statistical methods to optimize warehousing space, avoid stock outs saving millions in overstock & logistics costs in repackaging, pick-packing & holding costs Gave cross functional visibility to Channel Planners, Material Planners, and Field operations to track key performance indicators (KPI's) by integrating Python analysis to a Tableau dashboard resulting in 23% excess inventory reduction The University of Texas at Dallas · Graduate Teaching Assistant 01/2020 - 05/2020 Mentored 160 students through 1:1 coaching and delivering course lectures for courses Statistics & Data Analytics and Machine Learning in topics like hypothesis testing, ANOVA, Chi-Squared, sampling, A/B testing, quantitative analysis, etc I-link Infosoft Solutions Pvt Ltd. · Data Scientist 05/2018 - 08/2019 Developed new strategies to increase open rate of marketing emails by A/B testing different features from customer’s clickstream resulting in 11% more emails opened and an enhanced customer association of 19% to company's website Consolidated, cleaned and structured data from various historic ad campaigns and extracted key features to build a Logistic Regression using R with 92% accuracy to classify an email as spam Composed python and SQL code scripts to automate parallel email sending across different servers slashing 14 rework man-hours every week and boosting reach to 1.21 million potential customers Infosys Ltd. · Machine Learning Engineer 05/2017 - 04/2018 Developed, tested & incorporated new functionalities in Java to reorder previous purchases, send discount offers, and integrate PayPal/Google pay API's to augment new user adoption rates by 15% & increase code coverage by 13% Built a recommendation system using clustering algorithms by incorporating collaborative filtering and content filtering to show consumer-centric products to add to cart

Prioritized products to be displayed on clients platform by building a ranking matrix from user purchase histories and content similarity to increase online sales by 7%


Predict Appliance Energy Consumption

Implemented different dimensionality reduction algorithms Random Forest, forward selection, backward elimination, PCA & ICA Developed regression and gradient decent algorithms from scratch by minimizing cost functions for different learning rates Support Vector Machines - Classified house consumption for various gamma & kernels (Linear, Polynomial, Gaussian, Sigmoid) K-means clustering - Implemented Elbow method to identify the no of clusters and predicted consumption for households Decision Tree - Explored different decision criteria (Entropy, Information gain, Gini index, Chi-Square) to observe change in accuracies and implemented tree pruning to avoid overfitting Artificial Neural Networks - Experimented with different activation functions (Tanh, Sigmoid, ReLU), various optimizer (Adam, SGD, AdaDelta), no of hidden layers and different number of neurons Using News to Predict Stock Movements Techniques – XGBoost, CatBoost Aggregated news & market data to anticipate performance of a stock by apt selection of features from sea of available attributes Object Detection with OpenCV Techniques- Convolutional Neural Network, MobileNet Collected images from internet and labelled data to train a CNN to build a face mask detector with 96% accuracy using Keras, Tensor flow and OpenCV. Integrated the model to a live video camera to be used in real-time applications which require face- mask detection for safety purposes due to the outbreak of Covid-19 EDUCATION

The University of Texas at Dallas 01/2019 - 05/2021 M.S. Business Analytics - Data Science, GPA :- 3.74/4.00 Nirma University 06/2013 - 05/2017

B.S. Electrical Engineering


LANGUAGES & DATABASES Python, R, Scala, SAS, Java, Perl, C#, C++, Stata, MySQL, MS SQL, MongoDB, BigQuery BIG DATA & VISUALIZATION Hadoop, Spark, Hive, Sqoop, Impala, MapReduce, Tableau, PowerBI, Amazon Web Services FRAMEWORKS NumPy, Pandas, Scikit-learn, PyTorch, Keras, TensorFlow, Scipy, Matplotlib, Seaborn

Contact this candidate