AI for Oil & Gas Hackathon
November 10–14, 2025 — In person at CBPF, Rio de Janeiro
Final pitch: November 20, 2025 — SBGF 2025
25 participants
spots filled
Closed
23/09 a 31/10/2025 · R$ 100,00
5 days
November 10–14, 2025
About the event
The Hackathon is an applied innovation marathon that challenges multidisciplinary teams to develop Artificial Intelligence models for simulated data representative of the Oil & Gas sector.
Solutions must combine technical rigor, creativity, and industrial applicability, with final evaluation presented during SBGF 2025.
Objectives
Develop solutions for the analysis and interpretation of geophysical, petrographic, and NMR data with real application potential.
Integrate academia, industry, and research centers, strengthening collaboration and knowledge transfer.
Encourage the strategic use of AI in complex sector problems, from exploration to reservoir characterization.
Create a collaborative environment for networking and multidisciplinary partnerships.
Who can participate
Researchers from universities, research centers, and companies.
Professionals from data science, artificial intelligence, physics, geophysics, geology, engineering, and related fields.
Undergraduate students in any field related to the event's theme.
Graduate students (master's, PhD, or specialization).
Other professionals interested in AI applications in the oil and gas sector, even without prior experience in the field.
Participation Requirements
Have basic knowledge of programming and/or data science tools (Python, AI libraries, etc.).
Have your own laptop for use during the event.
Commit to attending in person throughout the hackathon days (November 10–14, 2025).
Agree to the event rules and the data usage policies.
Be willing to work in a team and collaborate in a multidisciplinary environment.
Limited spots: 25 participants.
Challenge Tracks
During the event, teams must choose two challenges organized into tracks:
RMN
Nuclear Magnetic Resonance
Task: Regression
Input: M(t) relaxation curves
Output: Oil/water fraction
Data: Simulated curves w/ noise (SNR)
Metrics: RMSE + R²
Thin Sections
Petrographic Thin Sections
Task: Image classification
Input: 256×256 synthetic RGB images
Output: spherulite, stromatolite…
Data: Synthetic dataset
Metrics: Accuracy and macro F1-score
Seismic
Seismic Facies Segmentation
Task: Semantic segmentation
Input: 224×224 grayscale patches
Output: Pixel-by-pixel masks
Data: NZPM seismic volume (CC-BY-SA 4.0)
Metric: IoU
Data Structure
Each track will provide two data subsets: training and test.
The training set will contain the labels needed to develop and fine-tune solutions.
The test set will be provided without labels and used exclusively for model evaluation.
(All datasets will be simulated and prepared for the event.)
How to Participate
Registration
Register through the official form and wait for email confirmation.
Schedule
Participate in activities according to the established schedule, including preparatory online lectures.
Attendance
Attend CBPF on the in-person days (8:30am–5:30pm). Address: Rua Dr. Xavier Sigaud, 150 — Espaço Oliveira Castro (Ground Floor).
Team Formation
Teams will be formed on the first day of the hackathon (November 10, 2025).
Development
Develop solutions as a team during the event days, with support from mentors and HPC infrastructure.
Final Submission
Submit the final solution (CSV/ZIP), commented notebook, and presentation slides by November 14, 2025.
Schedule
05–07/11
Preparatory online lectures, presenting the challenges, data, HPC usage, and space for participant introductions.
10/11
Official hackathon opening and team formation (in person at CBPF).
11–13/11
Project development, technical mentoring, and submission of partial results.
14/11
Final submission of solutions.
20/11
Final Session — SBGf 2025, 3pm to 5pm — BGP Geo Future Room (Mezzanine), Expo Mag, Rio de Janeiro. Address: R. Beatriz Larragoiti Lucas, s/n — Cidade Nova, Rio de Janeiro — RJ, 20211-175.
Judging Panel
Bernardo Machado de Oliveira Fraga
CBPF
Researcher at CBPF, working on AI and deep learning in geosciences for the Oil and Gas sector. PhD in Relativistic Astrophysics from Sapienza Università di Roma — Erasmus Mundus IRAP-PhD (2014). Develops deep learning and seismic inversion (FWI) for velocity model reconstruction and seismic imaging on HPC.
→ lattes.cnpq.br/8508146284928951
Pedro Barros Cotta Pesce
Petrobras
Senior Geophysicist at Petrobras, consultant at the intersection of education, machine learning, and geosciences applied to Oil and Gas E&P. Graduated and holds a master's degree in Physics from UFMG, dedicated for 15 years to corporate education through Universidade Petrobras.
→ www.linkedin.com/in/pedro-pesce-37a45233
Érica Kato Pacheco Ferraz
Petrobras
Geologist at Petrobras with extensive experience in petrophysics, reservoir characterization, and formation evaluation in the Campos and Santos basins. Graduated in Geology from UNESP, with specializations in Petroleum Geology (UERJ) and Advanced Petrophysics (University of Texas at Austin).
Bruno dos Santos Silva
Observatório Nacional
Researcher at ON/MCTI, computational geophysics applied to the Oil and Gas sector. PhD in Geophysics from UFPA, with experience in numerical modeling, seismic inversion (FWI), and subsurface imaging. Participates in R&D projects with Petrobras and TotalEnergies.
→ linkedin.com/in/bruno-dos-santos-silva-1b2b47a4
Thiago Freitas Lopes Conceicao
Petrobras
Senior Geophysicist at Petrobras, Sector Manager for Integration and Data Science for exploration geology. Bachelor's in Geophysics (UFBA), Master's in Civil Engineering, and MBA in Software Engineering (INFNET). Leads the development of digital products and analytics for exploration.
Project Evaluation
Solutions will be evaluated by the Judging Panel, composed of designated experts, considering the objective technical results presented as well as the qualitative criteria established in these rules.
Intermediate Publication (Technical Results by Track)
During the hackathon, each group will participate in two tracks. Results achieved in each track will be published on November 12 and 13 on a page on the event's official website, allowing everyone to follow the teams' progress.
Objective:
Provide transparency, encourage collective learning, and allow groups to see how others are progressing.
Format:
Each group will have its results displayed within the technical metrics of the chosen track, without ranking or classification.
NMR: RMSE and R².
Petrographic Thin Sections: Confusion matrix, Accuracy, and F1-score.
Seismic: IoU and related segmentation metrics.
This publication will be purely informative, serving as a way to exchange experiences and provide a healthy comparison among groups working on the same track.
On the last day of the hackathon, these results will not be publicly disclosed. They will be sent only to the judging panel as input for the final evaluation.
Final Evaluation by the Panel
The final decision rests exclusively with the Judging Panel, made up of experts from CBPF, Petrobras, and guests. Projects will be qualitatively analyzed according to the following criteria:
Accuracy and Technical Performance
Quality of results and robustness of the model.
Generalization Ability
The model's potential to extrapolate to new data.
Consistency of Approach
Methodological coherence and technical justification.
Innovation
Originality of the solution and creativity in the use of AI.
Applicability
Relevance and practical feasibility in the oil and gas sector.
Clarity of Presentation
Clear communication of methodology, results, and conclusions.
Teamwork
Collaboration, multidisciplinary integration, and task division.
Documentation and Reproducibility
Quality of records and possibility of replication.
Announcement of Results
The intermediate technical results page will be available only during the hackathon, showing performance by track without an overall ranking.
The official result will be determined by the panel after analyzing the solutions and announced later on the event website.
The panel may also award honorable mentions (Innovation, Applicability, Reproducibility) and recommend projects for continuation in future initiatives.
Prizes
🥇
1st Place
Petrobras Institutional Kit
Book "Seismic Signal Analysis"
Reservoir Engineering Book
"Best Solution 2025" Certificate
🥈
2nd Place
Petrobras Institutional Kit
Book "Seismic Signal Analysis"
"2nd Place 2025" Certificate
🥉
3rd Place
Petrobras Institutional Kit
Reservoir Engineering Book
"3rd Place 2025" Certificate
🎖️
Participants
Petrobras Institutional Kit
Hackathon 2025 Participation Certificate
Infrastructure
📍 Available at LAB-IA/CBPF:
HPC Server
Equipped with 8 NVIDIA Ada 6000 GPUs.
Resource Distribution
Each of the 5 teams (25 participants total) will have 1 dedicated GPU. The remaining 3 GPUs will be available on demand, upon justified request.
Pre-configured Environments
Docker, Conda, and Singularity, ready for immediate use.
User Accounts
Already created for each team, ensuring individual and secure access to the work environment.
Datasets
All participants will have access to the same initial dataset, organized in the teams' directories.
Additional Resources
Other LAB-IA infrastructure may be made available if the team demonstrates the need during development.
Registration
Period
23/09/2025 a 31/10/2025
Fee
R$ 100,00
Registration form
Registration closed
Includes (for all participants)
Lunch at CBPF during the in-person hackathon days
Morning and afternoon coffee (with snacks)
Rules
Accepting the data usage terms is mandatory (checkbox on the site at registration).
All data provided is simulated and synthetic, prepared exclusively for this competition as part of CBPF research projects, and is in the process of academic publication.
The models developed are the property of the participating teams.
Use of the datasets is restricted to the hackathon; redistribution or use outside the event context is prohibited.
The notebooks and code submitted (final commented submission) remain the property of the authors, but must be made available to the Organizing Committee for evaluation and record purposes. After the event, authors may choose to make the code public (open-source) or keep it restricted, accessible only for internal evaluation. The Organizing Committee guarantees it will not disclose the code without the authors' authorization.
Use of open-source libraries is permitted.
Each team must mandatorily deliver:
Final submission (CSV/ZIP).
Final commented notebook.
Presentation slides.
Organization
Event Co-Chairs
Marcelo Portes de Albuquerque (CBPF)
Bernardo Coutinho Camilo dos Santos (Petrobras)
Organizing Committee
Ana Paula Muller (Petrobras)
Clécio Roque De Bom (CBPF)
Elisângela Lopes de Faria (CBPF/FACC)
Márcio Portes de Albuquerque (CBPF)
Pablo Machado Barros (Petrobras)
Pedro Pesce (Petrobras)
Thais Fernandes de Matos (Petrobras)
Hackathon Mentor Team
Matheus Klatt
CBPF / FACC
Master's student in Computer Systems at COPPE/UFRJ, Bachelor's in Geophysics from UFF. FACC researcher on a joint project between CBPF and Petrobras. Works on developing deep learning applications for seismic data processing and interpretation, with experience in numerical modeling of elastic waves and reverse time migration. Was a Science Without Borders scholarship recipient in Germany.
→ lattes.cnpq.br/5273517356145686
Rayan Tadeu
CBPF / FACC
Master's student in Scientific Instrumentation at CBPF, graduated in Geophysics from UFF. Experience developing computer vision and machine learning algorithms applied to the analysis of petrographic thin sections, seismic data, and NMR. Worked on projects involving GANs, CNNs, and PINNs for geological data interpretation. Proficient in Python, PyTorch, TensorFlow, OpenCV.
→ lattes.cnpq.br/0043778647690124
Rômulo Rodrigues
CBPF / FACC
Master's (2022) and bachelor's (2019) in Geophysics from UFF. FACC fellow, working on applying AI and deep learning in geosciences for the Oil and Gas sector. Develops neural network techniques applied to computer vision for lithology detection and segmentation of petrographic features, such as porosity and dominant macroporosity.
→ lattes.cnpq.br/6769339563317613
Paulo Russano
CBPF
Master's student in Scientific Instrumentation at CBPF and technologist at the institution. Graduated in Information Systems. Works on the development and administration of HPC systems focused on AI, Machine Learning, and Deep Learning. Has experience with hardware, Linux servers, high-performance networks, and data center infrastructure, including multi-CPU and multi-GPU clusters.
→ lattes.cnpq.br/2130375444190292
Gabriel Gama
CBPF — LabIA
Physicist (UERJ/CBPF), postdoc at LabIA/CBPF working on LLMs and agents. Former project lead at Setup Automação, integrating YOLO + LabVIEW + ToF cameras for real-time inspection. Works in Microfabrication/MEMS, semiconductors, metrology, and automation with computer vision.
→ lattes.cnpq.br/2447429728456042
Luciana Olivia Dias
CBPF
Master's in Instrumental Physics from CBPF, working in the oil and gas sector since 2013. Developer of AI-based solutions including quantum simulation networks for demands involving well log, seismic, and NMR data. Works on several projects with Petrobras in partnership with CBPF.
→ lattes.cnpq.br/7347344568445543
Registration for the 2025 AI Hackathon is closed.
All spots have been filled.