Oleg Ovcharenko, PhD

Oleg Ovcharenko, PhD

AI & Energy expert at NVIDIA

AI & Energy Expert at NVIDIA

I spearhead the development and implementation of AI solutions across the energy sector, specializing in industrial agentic AI systems and domain-adaptive Large Language Models that transform operations in oil and gas and renewable energy industries.

With a strong research foundation and industry experience, I drive end-to-end AI projects that bridge the gap between theoretical advancements and real-world impact—from building HPC CUDA applications for exploration geophysics to deploying generative AI solutions that enhance decision-making and accelerate the energy transition.

News

Recent research

  • Multi-task learning for low-frequency extrapolation and elastic model building from seismic data

    Acquiring low-frequency data is challenging in practice for active seismic surveys. We propose to jointly reconstruct low-frequency data and a smooth background subsurface model within a multi-task deep learning framework. IEEE Transactions on Geoscience and Remote Sensing, 2022 Multi-task learning framework diagram showing low-frequency extrapolation and elastic model building

  • Direct domain adaptation through reciprocal linear transformations

    We proposed a general method for domain adaptation which blends features from source and target datasets by a set of linear operations such a cross-correlation and convolution. We explore the viability of the approach by transferring knowledge from training on gray-scale MNIST dataset to colored MNIST-M dataset. Frontiers of Artificial Intelligence, 2022 Direct domain adaptation visualization showing transformation from MNIST to MNIST-M dataset

  • Deep Learning for Seismic Data Reconstruction: Opportunities and Challenges

    Research project from my internship in CGG, Crawley, UK. Under supervision of Song Hou I explored benefits and pitfalls of using GANs for seismic data reconstruction. In the nutsell, high perceptual realism of reconstructed seismic data is not sufficient for real-world application. Physics also should be constrained. First EAGE Digitalization Conference and Exhibition, Nov 2020 Seismic data reconstruction using GANs showing original and reconstructed data comparison

  • Extrapolating low-frequency prestack land data with deep learning

    In synthetic framework we explore capability of a simple neural network to reconstruct low-frequency components of land data. SEG Technical Program Expanded Abstracts 2020 Neural network architecture for low-frequency extrapolation of prestack land seismic data

  • Style transfer for generation of realistically textured subsurface models

    We apply an iterative style transfer approach from image processing to produce realistically textured subsurface models based on synthetic prior models. These realistically textured models to be used in training datasets for machine learning applications in geophysics. Presented in Sep 2019, on SEG Annual meeting in San Antonio, USA Style transfer application showing synthetic and realistically textured subsurface geological models