An AI System for Institutional Trading Desks
We’re launching Axis: An AI Research Platform for Trading Desks. We’re starting with Commodities Markets.
About Us:
We both worked in quantitative trading across multiple firms. We learned a lot. It’s showed us how varied “trading” can be and how markets are best approached.
About Axis:
We started building Axis on the belief that models could be incredibly powerful when applied to traded markets.
This mainly came from our experience as traders. Trading is highly context driven, yet the majority of qualitative market information goes unparsed: real-world news, broker data, macro narratives, and more. This information is high-signal to a trader but unstructured. For the first time, LLMs provide an effective way of analyzing this information across multiple sources.
In Short: We believe models reveal what a trading system could be: not a data feed, but a tool that actively enhances a trader’s view of the market.
Our first market: Commodities
Commodities are real-world markets that make up the movement of energy, agricultural products, metals, chips, and more around the globe. They make up roughly a third of global goods exports.
Commodities trading is a unique domain. Like equities, there are futures/options and other publicly traded derivatives that track commodities at the macro scale. This is where banks and hedge funds typically trade.
However, there is an entire separate group of “physical” commodities traders who are responsible for moving commodities (oil, grain, copper) to locations around the globe. Their “trading” is the original form of “trade”: transporting materials to another location, taking on storage and transportation risk, receiving a premium for the service. Oil companies like Shell and Chevron have large commodities trading units responsible for trading/hedging their supply.
These traders still operate like normal trading desks, hedging and speculating in public markets, but they are also responsible for deciding storage and transportation. They track the markets at a granular and varied scale: transportation logistics, regional demand/inventory, derivatives positioning, macro narratives, and more.
Engaging in these markets is complicated and it’s where we think our models shine. Axis allows a commodities desk to deploy custom models for monitoring their strategy 24/7.
How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?)
We’ve known each other since pre-school (as we’re told) and both worked in quantitative trading across different firms. We left our jobs on the core belief that models would fundamentally change how we interact with markets. While today’s trading systems are effectively data feeds with premade workflows, reasoning models suggest that trading systems can be something much more: tools that thoughtfully monitor and react to markets.
