Dr. Ivan Abakumov

Lead Data Scientist at HQ revenue
  • Claim this Profile
Contact Information
us****@****om
(386) 825-5501
Location
Berlin, Berlin, Germany, DE
Languages
  • Russian Native or bilingual proficiency
  • English Professional working proficiency
  • German Elementary proficiency

Topline Score

Topline score feature will be out soon.

Bio

Generated by
Topline AI

You need to have a working account to view this content.
You need to have a working account to view this content.

Credentials

  • Natural Language Processing
    Coursera
    Jul, 2021
    - Nov, 2024
  • Deep Learning Specialization
    Coursera
    Jul, 2019
    - Nov, 2024
  • Sequence Models
    Coursera
    Jul, 2019
    - Nov, 2024
  • Neural Networks and Deep Learning
    Coursera
    Jun, 2019
    - Nov, 2024
  • Convolutional Neural Networks
    Coursera
    Jun, 2019
    - Nov, 2024
  • Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
    Coursera
    Jun, 2019
    - Nov, 2024
  • Structuring Machine Learning Projects
    Coursera
    Jun, 2019
    - Nov, 2024
  • Unconventional Reservoir Geomechanics
    Stanford Online High School
    Jun, 2019
    - Nov, 2024

Experience

    • Germany
    • IT Services and IT Consulting
    • 1 - 100 Employee
    • Lead Data Scientist
      • Apr 2020 - Present

      • Plan, organize, and monitor AI-based projects in the company • Manage work of the Data Science team • Interview and onboard new employees • Provide advanced analytics for the Marketing & Business development teams • QC of the data from external data providers • Report results to stakeholders & auditors • Consult principle customers on AI-based questions • Publicly announced highlights of the Data Science team under my supervision: - daily forecast city demand for 1400 destinations - daily estimate the yearly hotel demand for 700.000 properties (more than 58% worldwide) - daily analyze 90 million rates - in real-time provide occupancy forecast for customers Show less

    • Germany
    • Non-profit Organizations
    • 1 - 100 Employee
    • Instructor
      • Mar 2020 - Oct 2023

      • I prepare and keep the DGMK advanced training course "Machine Learning for Subsurface” (3-day intensive course for industry professionals). • Next course is planned for March 2023 • I prepare and keep the DGMK advanced training course "Machine Learning for Subsurface” (3-day intensive course for industry professionals). • Next course is planned for March 2023

    • Germany
    • Research Services
    • 700 & Above Employee
    • Postdoctoral Researcher
      • Sep 2017 - Apr 2020

      • Developed innovative approaches for detection and localization of microseismic events recorded simultaneously by the surface and borehole arrays • Delivered physically based analytical expressions for arrival-time uncertainties • Initiated and developed departmental educational activities including master student courses “Inverse problems and machine learning for exploration geophysics” and “Deep learning for exploration geophysics” • Developed innovative approaches for detection and localization of microseismic events recorded simultaneously by the surface and borehole arrays • Delivered physically based analytical expressions for arrival-time uncertainties • Initiated and developed departmental educational activities including master student courses “Inverse problems and machine learning for exploration geophysics” and “Deep learning for exploration geophysics”

    • Germany
    • Research Services
    • 700 & Above Employee
    • Research Assistant
      • Oct 2013 - Aug 2017

      Multidimensional stacking • Performed systematic analysis of stacking operators; extended the theory of DSR-based stacking operators to the case of converted waves, the case of anisotropic media and the 3D case • Delivered numerous software products: 3D i-CRS code, CRS-based prestack data enhancement and interpolation code for arbitrary acquisition, PreSDM code Time-lapse seismic • Developed an algorithm of advanced cross-well velocity analysis • Applied the algorithm to time-lapse cross-well data acquired in South Oman to track steam injection • Achieved the improved anomaly focusing compared to the time-lapse tomography result • Developed 4D FWI algorithm based on the pseudospectral Fourie method to further improve the anomaly focusing Show less

    • United Kingdom
    • Oil and Gas
    • 700 & Above Employee
    • Trainee
      • Aug 2014 - Nov 2014

      • Implemented the original method of time-lapse shallow velocity anomalies reconstruction by inverting virtual diving waves extracted from 4D OBN data • Delivered a new QC method for node-related static corrections • Implemented the original method of time-lapse shallow velocity anomalies reconstruction by inverting virtual diving waves extracted from 4D OBN data • Delivered a new QC method for node-related static corrections

    • Russian Federation
    • Higher Education
    • 200 - 300 Employee
    • Engineer
      • Jan 2011 - Oct 2013

    • Czechia
    • Oil and Gas
    • 1 - 100 Employee
    • Trainee
      • Feb 2013 - May 2013

      • Developed 3D VTI eikonal solver based on the fast sweeping method • Integrated the eikonal solver into the Seismik's software (significantly speeded up the localization procedure and allowed to account for 3D velocity models) • Developed 3D VTI eikonal solver based on the fast sweeping method • Integrated the eikonal solver into the Seismik's software (significantly speeded up the localization procedure and allowed to account for 3D velocity models)

    • Germany
    • Research Services
    • 700 & Above Employee
    • Visiting Researcher
      • Feb 2012 - Apr 2012

      Applied the method of multidimensional stacking of converted waves to realistic synthetic and field data Applied the method of multidimensional stacking of converted waves to realistic synthetic and field data

    • Germany
    • Research Services
    • 700 & Above Employee
    • Exchange Student
      • Oct 2011 - Dec 2011

      Compared the analytical reflection coefficients at a thin fluid layer with the numerical results obtained by means of wavenumber integration technique (OASES package) and finite-difference method Compared the analytical reflection coefficients at a thin fluid layer with the numerical results obtained by means of wavenumber integration technique (OASES package) and finite-difference method

    • Germany
    • Research Services
    • 700 & Above Employee
    • Visiting Researcher
      • Aug 2011 - Aug 2011

      • Implemented the double-square-root traveltime approximation for converted waves into the CRS code • Proposed gamma-CMP gathers and a pragmatic search of stacking parameters for converted waves • Implemented the double-square-root traveltime approximation for converted waves into the CRS code • Proposed gamma-CMP gathers and a pragmatic search of stacking parameters for converted waves

    • Germany
    • Research Services
    • 700 & Above Employee
    • Exchange Student
      • Nov 2010 - Nov 2010

      Derived the reflection and transmission coefficients for plane waves incident on a thin fluid layer (a model of a hydraulic fracture) Derived the reflection and transmission coefficients for plane waves incident on a thin fluid layer (a model of a hydraulic fracture)

Education

  • Universität Hamburg
    Doctor of Philosophy (PhD), Geophysics
    2013 - 2017
  • Saint Petersburg State University
    Master of Science (M.S.), Physics
    2009 - 2012
  • Saint Petersburg State University
    Bachelor of Science (B.S.), Physics
    2005 - 2009

Community

You need to have a working account to view this content. Click here to join now