Eight years inside ExxonMobil — from compliance to product

Starting as an Environmental Advisor managing 40+ state and federal regulations, moving through process engineering, and ultimately leading risk strategy for a team of engineers and inspectors. The foundation for everything that followed.

Starting in compliance

I joined ExxonMobil in 2006 as an Environmental Advisor — a role that put me at the intersection of operations, legal, and regulatory compliance from day one. My job was to advise operations teams while managing compliance across more than 40 state and federal environmental regulations across multiple facilities.

In practice, this meant being the person in the room who understood both what the regulation actually required and what the operational reality was. That tension — between what the rule says and what's actually feasible in a complex industrial environment — became one of the most useful things I've ever learned. It's the same tension I navigate in AI today: what the technology can theoretically do versus what it can reliably do in a real production environment.

Process engineering — learning how the system actually works

From 2009 to 2010 I was a Process Engineer running naphtha hydrotreater and reformer operations. This was hands-on work — managing the unit economics of refinery operations, working with microeconomists to match production to demand signals, and supplying Southwestern US markets with high-octane gasoline blends.

Process engineering at a major refinery teaches you something that's surprisingly rare in tech: how to think about systems under constraint. You're not optimizing for a single variable. You're managing temperature, pressure, catalyst performance, feedstock quality, market price, and maintenance windows simultaneously — in real time, with real consequences for getting it wrong. That systems thinking carried directly into how I build and evaluate AI systems today.

Product development — first taste of PM work

As Product Development Manager from 2011 to 2012, I led sub-$4M projects as the primary PM and contributed to sub-$20M projects as a junior PM. This was my first formal exposure to the product management discipline — defining scope, managing stakeholders, tracking delivery, and being accountable for outcomes I didn't fully control.

The projects at this scale in a company like ExxonMobil involve more organizational complexity than almost anything in tech. Getting alignment across engineering, operations, legal, procurement, and finance — on a project that will run for 18 months and has no tolerance for scope creep — is an exercise in clear communication and constant prioritization. Those skills transfer completely.

Risk strategy — leading people and systems

My final role at ExxonMobil was Risk Strategy Manager — leading a team of engineers and inspectors responsible for asset risk assessment across ExxonMobil's infrastructure. The work was about making decisions under uncertainty: which assets pose the highest risk, which inspections are most likely to surface problems, how to allocate limited inspection capacity across a large and complex asset portfolio.

Sound familiar? It's essentially the same problem as predictive maintenance AI — but solved with human expertise, structured frameworks, and a lot of institutional knowledge that lived in people's heads rather than in data systems. That experience gave me a visceral understanding of why AI is hard in industrial settings: not because the math is complicated, but because the knowledge that would make the AI useful is fragmented, implicit, and hard to capture.

Why this matters for my consulting work

The eight years I spent at ExxonMobil gave me something I couldn't have gotten any other way: genuine fluency in how large industrial operations actually work. Not the sanitized version that shows up in vendor presentations or analyst reports — but the version where the regulations conflict with each other, the data is in four different systems, and the most important knowledge is held by a 25-year veteran who's retiring next year.

When I consult with oil and gas companies on AI strategy today, I'm not learning the domain on the job. I'm applying 19 years of accumulated context — including 8 years of living inside the company that many of my clients aspire to. That changes the conversation.