 As geoscientists, understanding Earth's subsurface is the primary challenge. We spend days, months, sometimes years interpreting complex data and images that give us hints about the rock below our feet. Every bit of new information is a step towards a better decision, and it's not easy to get it right. It's common to have to drill around seven exploratory wells until we discover one new source. That's an exploratory success rate from 10% to 15% in new frontiers. Drilling a single well in the pre-salt area, for example, can cost anywhere between $100 and $300 million U.S. dollars. But now, with the help of AI, we have a cognitive advisor that's changing the way we investigate the subsurface, building the link between different knowledge sources, experts, and technologies. With the aim of improving success rates and reducing risks and costs, tests already indicate a significant performance improvement of a specialist working on the process of interpreting seismic images. And by using the prototype to predict rock properties, we've already obtained encouraging results. And we're just starting. This solution is the result of a research effort between GALP and the IBM Research Lab to empower geoscientists with information that matters, where it matters, for the maximum impact. This AI-based advisor applies visual comprehension techniques to create intelligent analyses of seismic 3D images, delivering important insights, and accelerating mentally heavy tasks. To quickly identify geological structures that could yield oil and gas, along with powerful AI algorithms, the system integrates information from different sources. Data of available wells, reviews of previous studies, contextual data, and user's annotations. Every input is registered and combined in a unique environment. They create a seamless foundation for multi-user collaboration, improving the overall knowledge about the target prospects areas. During our exploration journey, we may test several hypotheses and review previous decisions as we gather new, significant data. With an accurate geological characterization, we have a faster and more structured risk assessment. This AI technology is poised to enhance and even reinvent our decision-making process in geological exploration, allowing the discovery of hidden connections that help unlock the full value of data. This advisor is our most accessible peer, assisting with data awareness and reasoning that allow us to gain a better and faster understanding of the subsurface. An environment of continuous learning and evolution is created, helping us face the growing challenges of the oil and gas industry.