IGI's Dan Cornford will be attending and presenting at IMOG 2025 in Porto 8-11th September.

He will be talking about the challenges of applying our new machine learning tools to geochemical data, motivated by work with DNO ASA on producing a holistic understanding the oils and source rocks of the North Sea.

The talk will focus on where machine learning can help, and where we still need geochemical expertise. Using examples, we will show that at least for the foreseeable future machine learning will augment, not replace conventional geochemical and geological expertise. This can be largely explained by the complexity of most reservoir oils, which are  almost always mixtures and affect by many alteration and charge characteristics, meaning in reality every reservoir oil is pretty much unique.