- Current Company 2018 - Present
The designer and developer of Predictive *** platform at Current Company The platform is implemented and deployed where customers can find when will the next inspection be required and what will be the error rate. The *** platform enables customer for the -Prediction for the next inspection using machine learning/deep learning Other responsibilities are: -Build machine learning solutions and Proof of Concepts using the best combination of existing and open source technology and *** automotive predictive capabilities. -Leading technical demonstrations of *** both for technical and non-technical audiences. -Mastering technical tools for manipulation of data sets (R, Python). -Use data visualization techniques with open source libraries (R, Python). -Explore the data by searching for relationships, trends, anomalies, and outliers. -Modify the data by creating, selecting, and transforming the variables for model selection. -Model data by using analytical tools to train statistical/probabilistic models to reliably predict desired outcomes. -Develop mathematical models for detecting fault of Wind Turbine Blades.
- Gram&Juhl 2017 - 2018
Develop machine learning algorithms in Turbine Condition Monitoring System (TCM) in order to: • Identify the existence and location of damage in wind turbines • Identify the type and severity of damage in wind turbines • Modeling state of the art Condition Indicators • Development of new Condition Indicators • Prediction of Health Indicator for some components of Wind Turbines • Identify the existence and location of damage of the components • Prediction of Remaining Useful Life of the components of Wind Turbines/ the components (of the Wind Turbines) • Data Validation • Collaborating with the data engineering team on developing data pipelines • The prediction error and reliability must be improved to cope with the market requirements
- Presidents Institute 2017 - 2017
Responsible for: -Automatic lead generation -Lead scoring -Predictive analysis of customer behavior -Segmentation and Recommendation System