Qiran Low: Commodities and AI

Hey, it's Qiran. Nice to meet you!
I started my career at Goldman Sachs in commodities sales and trading, covering clients through some of the most volatile years in recent market history (Covid, and then the Russian-Ukraine war). After that, I became the sole thermal coal specialist at a commodities hedge fund and built its thermal coal analytics capability from the ground up: including global supply–demand balances, seaborne flow models, vendor revision trackers, and various dashboards and scenario tools. Most recently I had a short stint in physical trading where I learnt how physical contracts, negotiation, operations and logistics work.
My background is in commodities, so the content here naturally leans that way. But I hope this blog can still be interesting for anyone curious. So I'll keep my language simple and explain key commodities concepts along the way. No barriers to entry, no gatekeeping.
So far I've explored topics like how AI can support commodity trading and research from gas-demand models to AI-agent benchmarks. I'm also working on making AI play a cargo allocation game for physical trading allocations. Sometimes I do deep dives into topics which I find interesting, such as building a global compute cost curve and looking into USO ETF rolls.
What you’ll find here would be practical experiments, working tools, and honest assessments of what AI can and cannot do in commodities workflows.
I'm still figuring things out, so let's explore and learn together!