Workstream: ML/AI

12:30  Predicting Key Geothermal Reservoir Properties Using AI/ML

- Wojciech Panek, Engineer - Process Data Subsurface&Wells, Shell

A Case Study from Northern Croatia, showcasing the role of data in energy transition of E&P business.

13:30  Use of acoustic sand detectors and music processing techniques for mitigating chalk influxes 

- Dr. Ada Ortiz-Carbonell, Data Scientist, AkerBP, (Expert Analytics AS)

An unusual approach for an unusual challenge. Chalk reservoirs are rare, and as such pose problems that few have confronted before. We show how data, innovative sound processing techniques, acoustic tech, and pushing acoustic sensors beyond their capabilities can help mitigate the problematic well plugging caused by chalk influxes. While chalk can block production, we are fighting back with sound. 

14:30  High-Resolution Elastic Property Prediction and Fluid Detection with Deep Learning  - Åsmund Heir, CEO, RagnaRock

Can AI support fluid detection in field development? This presentation explores how machine learning complements traditional methods by predicting high-resolution elastic properties from seismic and well data, using rock physics to detect fluids with high accuracy in a real-world case study.

15:30  Real-time gas-oil ratio estimation in Pre-Salt offshore wells using data-driven regression models  - Leila Araujo Ribeiro Farias, Reservoir engineer, Petrobras

Enhancing GOR monitoring and optimizing production strategies through machine learning techniques utilizing pressure and temperature sensors in complex offshore environments.

16:30  Transforming Legacy Reports into Strategic Assets - Gen AI-Powered Digitalization of Legacy Geological and Well Test Reports into Geological Database for Enhanced Exploration and Production 

- Chafaa Badis, AI & Data Science Advisor, Halliburton

The upstream oil and gas industry faces challenges in extracting data from unstructured formats like PDFs, slowing decision-making in mature fields. To address this, INA and Halliburton Landmark developed an AI-driven workflow using OCR, NLP, and Generative AI to digitize and structure geological data. This reduced data processing time by 95%, improving efficiency, accuracy, and reservoir management. The solution empowers faster, data-driven decisions and supports sustainable production strategies.

08:30  The Anatomy of an Agent

- James Patton, Global Data Platform's Manager, AWS

This session explores how to design a domain-aware agent that enables natural language interactions and autonomous data discovery, calculation, and delivery.

09:30  Blueprints We Didn't Draw

- Michaël van der Haven, Vice President Consulting Expert, CGI
How the current state of AI can be used to organically grow a Data Mesh, create true Data Products and deliver value in weeks instead of months. And we'll show you how you can do this by yourself!

10:30  Not Brains vs. Algorithms—Both: Rethinking Seismic Interpretation

- Adnan Latif, VP of AI & Global Services, Bluware

This session will focus on explaining where we are in terms of technological readiness for fully transitioning from human cognition to automation in seismic interpretation. It will also share a few ideas on how, instead of replacing one (Geoscientist) with another (AI), we could combine human intelligence with artificial intelligence to leverage the best of both worlds in seismic interpretation.