Saeid Akhavan

EPC company in Metal, Oil, Gas and Petrochemical Industry.

• Data and feature engineering for data logs of more than 2000 sensors assembled on plant (shaft, reformer, scrubbers, and compressors.)

• Predicted metalization and carbon percentage of shaft furnace output with 85% accuracy.

• Designed many dashboards and visualizations to monitor and manage the plant performance.

Online supermarketfor more than 2500 stores all over the country.

• Main data analyst of Marketing and Product team.

• Performed a time series Shopping Cart Analysis of a dataset with more than 30k baskets per day to recognize the ordering pattern of the customers.

• Implemented an OMS dashboard with Power BI to predict the Cancellation Rate, Delivery Delays, Ticket Tracking and other Order Management KPIs.

Supervisor: Dr. Amin Zollanvari, Texas A&M University, USA.

• Developed a fault prognosis method to reduce maintenance costs of transformers in large scale energy companies.

• Examined the possibility of using deep neural networks (DNN) in capturing the hidden patterns of vibration time-series to predict the transformer under and over excitation and inter-turn fault progress prediction in early stages.

• Adopted Keras library and deep recurrent neural network (RNN) architecture including unidirectional and bidirectional gated recurrent units (GRUs) and long short-term memory (LSTM) models. The constructed RNN for predicting excitation voltage exhibited a remarkable performance with a relative absolute error (RAE) of 0.56%.