*

Milad Najafbeygi

Cell: 405-***-**** LinkedIn: https://www.linkedin.com/in/milad-n/ E-mail: ad2uqj@r.postjobfree.com SUMMARY

Expert Data Scientist with a strong focus on SQL, Python and Sisense in Aerospace and Oil. Excels in machine learning and predictive analytics, driving significant process improvements and cost savings. Committed to innovation and effective teamwork in fast-paced environments.

TECHNICAL SKILLS

Code

• SQL

• Python

• C++

• AMPL

• Fortran

Data Modeling

• Data Warehousing

• Dimensional modeling

• Data Transformation

Source Control

• Git

• GitLab

• Bitbucket

Machine Learning

• Random Forest

• Decision Tree

• Support Vector Machine

• Linear Regression

• Logistic Regression

• Neural Networks

• Natural Language

Processing (NLP)

Statistical Methods

• ANOVA

• Z score

• T-test

• P value

Analytics

• Pandas

• Numpy

• Sklearn

• R

• MATLAB

Relational Database

• Postgres

• MySQL

• Microsoft SQL Server

Data Visualization

• Sisense

• Matplotlib

WORK EXPERIENCE

Data Scientist – Revenue Management Solutions, LLC July 2022 – October 2023 Oklahoma City, OK

Identified and implemented an automated solution that resulted in a 72% reduction in errors and a 400% increase in monthly processed claims (from 1 million claims processed to 4 million claims).

• Created a training, testing and validation set of data by labeling the important components of the insurance claim template.

• Utilized SQL and Postgres for advanced anomaly detection in insurance claims, identifying key mismatches and significantly reducing fraudulent claims, enhancing payout accuracy and operational efficiency.

• Improved data accuracy by implementing data cleaning and NLP techniques, including the Jaccard-Winkler algorithm, achieving a 72% match rate in name normalization, reducing manual entry errors, and enhancing data integrity.

• Developed and optimized live data visualization dashboards in Sisense, streamlining the data pipeline to enable real-time monitoring, leading to quicker decision-making and enhanced data integrity management.

• Enhanced database performance by eliminating redundant data and applying common table expressions (CTE) in queries, resulting in more efficient data processing, faster access, and reduced storage overhead.

• Restructured existing queries to update business logic and improved data pipeline to reduce reporting downtime. Data Scientist – Schlumberger Global Energy Services & Equipment – Graduate Research Assistant January 2019 – July 2022 Oklahoma City, OK

Reduced testing costs by $24 million annually by developing a leak rate model using Argon instead of Helium. Company continues to use the model and apply it across along all lines of their business.

• This resulted in an annual cost and time savings of $24 million dollars for testing seals across every product in the company.

• Reduced the cost of acquisition for testing seals by utilizing Aragon instead of Helium resulting in 10x savings.

• Developed an SVR model to reduce the cost of testing seals in oil and gas products used across the company. 2

• Got buy-in from the EPA that Aragon is an acceptable substitute for testing leak rate instead of Helium based on my research.

• Reduces regulatory liability by reducing fines for pollutant exposure. University of Oklahoma - Master’s Capstone Project July 2019 – July 2022 Modeling the decision-making process of auto insurance policymakers to understand the biases in their decision-making process and data interpretation. Introduced a novel way to examine bias by employers from a data science perspective.

• Created a predictive model using SVR, linear regression and neural network to predict the premium of insurance policyholders.

• To improve the accuracy of the model hyperparameters tuning, gridsearchCV and cross-validation techniques have been used.

• Tuned hyperparameters to increase the model’s accuracy.

• Customer gender, marital status, and location have been considered as variables.

• Results show on average, a female who has been divorced has the highest premium value for the company, and it is almost 3.6% higher than non-gender and non-marital status customers on average. EDUCATION

M.Sc. in Data Science and Analytics, University of Oklahoma, 2022 M.Sc. in Mechanical Engineering, University of Oklahoma, 2022 M.Sc. in Aerodynamic-Aerospace Engineering, University of Tehran, 2014 B.Sc. in Mechanical Engineering, Urmia University, 2011 RESEARCH

• Thesis: Using SVR, created a predictive model based on experimental data “analyzing Argon emission on rotational shafts in V-rings seal.

• Thesis: Reduced-order model for the reconstruction of the flow field around 2D bodies based on a combined form of POD and HOSVD methods.

Contact this candidate