- Tensorflow/Keras
- Pytorch
- Scikit-learn
- Scikit-Image
- OpenCV
- OpenPNM
Sa'eed Telvari | Curriculum Vitae
Department of Petroleum Engineering
Amirkabir University of Technology, Tehran, Iran
Research Interests
- AI Applications in Subsurface
- Porous Media
- Image Processing
- Numerical Simulation
- CCUS
- Computer Vision
Education
- Amirkabir University of Technology, Tehran, Iran
- M.Sc. of PE - Reservoir Engineering
- GPA : 3.61/4
- 2022-present
- Amirkabir University of Technology, Tehran, Iran
- B.Sc. of Petroleum Engineering
- GPA : 17.43/20 (3.57/4)
- 2018-2022
Publications
- "Prediction of two-phase flow properties for digital sandstones using 3D convolutional neural networks"
Saeed Telvari, Mohammad Sayyafzadeh, Javad Siavashi, Mohammad Sharifi
(Link)
- "Fractured Rock Image Segmentation using Autoencoder-Based Deep Learning. (In progress)"
Saeed Telvari, Mohammad Sharifi
Internship
- Iranian Offshore Oil Company
- Aug-Sep 2021
Honors & Awards
- Ranked within the top 2% in Iranian University Entrance Exam for masters degrees
- Granted direct admission for graduate study from Talented Student Office of Amirkabir University of Technology
- Received national undergraduate scholarship (full tuition waiver)
- Ranked within the top 4% among more than 140,000 students in Iranian University Entrance Exam for bachelor degrees
- Recognized as a talented student in the entrance exam of NODET for middle and high school
Academic Projects
- Automatic interpretation of well tests
- Implemented LSTM to predict reservoir properties from pressure derivative and dp
- Developed a script to match analytical model to data
- Developing an extended local upscaling module in MRST
- Remodeled a local upscaling module with periodic boundary conditions into an extended local one
- Implemented and compared with local and fine scale pressure field
- Conjugate heat transfer between rock and fracture fluid using OpenFOAM v9
- Extracted and partitioned a micro-scale 3D image of a fractured granite rock
- Simulated injection of CO2 with a lower temperature into a fractured granite with high temperature
- Training machine learning models to identify formation lithology
- Used 10 conventional well log data as input and mud logging data as output
- Developed FNN (pytorch), Decision Tree (tensorflow), and clustering algorithms such as SVM and k-means
Skills
Courses & Certificates
- Python 3 Master Course
- Deep Learning: GANs and Variational Autoencoders
- The Basics of Transport Phenomena
- Machine Learning
- Deep Learning and Neural Networks
- TensorFlow for Deep Learning with Python
Language
- Persian
- English
- Native
- Professional Working Proficiency
TOEFL : 108- R : 30
- L : 28
- S : 22
- W : 28