The data analytic side of oil majors is extremely sophisticated and complex: these are groups dealing with massive sets of all-sensor geologic data and using them to make multi-billion-dollar bets as to what's going to happen thousands of feet under the earth under complex and immense temperature and pressure gradients.
Even outside of those departments, the challenges are significant (though companies underinvest in digital in those areas). Every downstream refinery, for example, is a unique beast full of custom equipment, sensors, and systems, so tying those into digital twins in a scalable, effective, and efficient way is a massively hard problem, and one that has potentially dire real-world consequences. Midstream distribution is full of complex optimization issues from both physical and financial points of view. And upstream E&P, in addition to the geology side mentioned above, is full of literal life-and-death decisions that require sophisticated real-time analytic models that can guide driller behavior.
Leaving aside the moral issues raised by working for carbon-intensive energy industries, the technical challenges are significant, impressive, and fascinating to work on -- more so, I would argue, than trying to slightly increase social media ad click rates.