Data Scientist

MALAY SHAH

918-***-**** ************@*******.*** www.linkedin.com/in/malay-shah-3302 https://github.com/mjs3302 SUMMARY

Result oriented data scientist with 4 years of experience leveraging the informational assets to drive operational efficiency with improved decision making by analyzing products and its KPIs

Certified in AWS Machine Learning with experience in predictive analytics and transforming data into insights WORK EXPERIENCE

Tulsa University Separation Technology Projects Tulsa, OK Research Assistant 2017- Present

Developed a predictive time series machine-learning model to predict equipment failure by preprocessing and modeling IoT sensor data to an accuracy of 96%.

Queried, and integrated data from multiple sources along with conducting exploratory risk and data analysis to eliminate outliers, identify trends, and eliminate noisy data using Python.

Developed end-to-end machine learning model was prepared to run in on-field production environments with an RSME of 3%, reducing energy wastage on fluid separation.

Prepared a failure analysis document on system failure with a recommendation for process improvement

Engaged with enterprise energy clients to make model modifications as per their requirements.

Mentored junior engineers with regards to the choice of optimal variables on modeling metrics and published thesis and paper on a predictive model

Tinita Engineering Pvt Ltd Mumbai, India

Data Scientist, Operations 2016 - 2017

Developed a business intelligence dashboard using SQL, Tableau for inventory management which was used to identify inefficiencies in welder performance and material wastage.

Performed cost-benefit analysis by developing a model for selecting an appropriate vendor and reliable supplier of subparts reducing outsourcing cost by $200k.

Defined and evaluated KPI’s for operations and supply chain planning to forecast delivery time of outsourced products with an accuracy of 0.5 days

Conducted operational, financial and mathematical analysis to support root cause analysis on manufacturing and dispatch delay reducing mean time to resolution (MTTR) by 73%

Automated monthly inventory requirement using VBA scripting, saving five hours per week EDUCATION & CERTIFICATION

MS (Thesis), Mechanical Engineering, University of Tulsa 2017 - 2019 Data Science Coursework: Machine Learning, Data Visualization, Decision Analytics & Modeling, Statistical Analysis BE, Mechanical Engineering, University of Mumbai 2012 - 2016 AWS (Amazon Web Services) Certified Machine Learning – Specialty SKILLS

Environment and Techniques: Python, SQL, MatLab, Tableau, Looker, R, SAS, JavaScript, C++, Power BI, Excel, GitHub, GIT, LabVIEW, Airflow, Jenkins, Docker, Kubernetes, AWS, Azure, Hive, Hadoop, Spark, Scala, REST API Data Science Skills: Regression, Classification, Time Series, Demand forecasting, ARIMA, LSTM, Predictive Modeling, Supply Chain Analysis, Natural Language Processing PROJECTS

1. Airlines Passenger Data:

Time Series Airlines passenger data was checked for stationarity with techniques like rolling plots, and ACF-PACF plots. 2. Auto MPG Prediction:

Developed a linear regression model on variables that affect vehicle mileage with Ridge and Lasso Regularization to reduce overfitting and error by 85%

3. Credit Card Fraud Detection:

Used AWS SageMaker’s Linear Learner to deal with class imbalance and tuning hyper parameters (precision and recall) to reducing payments fraud with an accuracy of 99.96%. 4. IMDB Sentiment Analysis:

NLP reviews modeled on a Random Forest classifier was deployed with a Lambda function on AWS with an accuracy of 93%

Contact this candidate