Why We Built Loqsea: AI in Commodity Trading Only Works on Live Data

A conversation with Reza Tareen, Managing Director at Loqsea Technology.

Every CTRM vendor now has an AI story. Far fewer will talk about what sits underneath it.

This is the sixth article in our Why We Built Loqsea series. Reza Tareen, Managing Director at Loqsea Technology, has spent the past year closer to both the product roadmap and the client conversations than almost anyone at the firm. This is his account of why Loqsea built Tommy the way it did, and where AI goes next in the platform.

The industry has moved quickly on AI over the last twelve months. What has moved more slowly, and matters considerably more to a trading desk, is the question of what these tools are actually reading.

Where has AI in commodity trading actually got to?

The question has changed.

A year ago, the debate across commodity trading was whether AI belonged in the workflow at all. That argument has largely resolved itself. What has replaced it is narrower, more practical, and harder to answer.

“A year ago most of the industry was still deciding whether to adopt AI in processes at all. That question has largely gone. What has taken its place is a much narrower conversation about what these tools are permitted to touch and who carries the responsibility when they get something wrong.”

For most firms, the constraint is no longer appetite.

“Most of the firms we speak to are keen to make use of AI but are still working out what they would genuinely trust with their own data. Having a partner like Loqsea in that conversation can considerably speed up the benefits of building the right AI ecosystem.”

Where is AI genuinely being used on a trading desk today?

Ahead of itself.

There is a wide gap between what the market is discussing and what is actually running on a commodity derivatives desk. The realistic applications all share one characteristic: a person is still in the loop.

“The realistic applications at present are the ones where a person still makes the decision and takes the action. Reaching a number faster. Checking a position without stopping what you are doing. Understanding exposure without waiting on someone else to produce it.”

The wider conversation has moved well beyond that, towards systems that act on their own. In commodity derivatives, the consequences of getting that wrong are different.

“The wider conversation has moved a long way ahead of that, towards systems that act independently, and in some markets that has already begun. On a commodity derivatives desk the consequences of an error are different, which is why adoption there will take longer, and why it probably should.”

Why did Loqsea build Tommy?

Removing the step between the question and the answer.

“The answer is already sitting in the platform, which hosts an enormous amount of data. The difficulty has always been reaching it. Someone wants their VaR, and the route is a report, or a request to operations, or an export into a spreadsheet. None of that is complicated work. It simply takes time, and it happens repeatedly through the day.”

Tommy is Loqsea’s AI chatbot for commodity trading analysis, built directly into Risk Manager. The problem it addresses is not a shortage of data. It is the distance between a trader and data they already own.

That repetition is the cost, and it is what Tommy was built to remove.

“Tommy is an AI chatbot for commodity trading analysis, and what it does is remove the step between having a question and having the answer. Our clients ask their own data directly, in the words they would have used to ask a colleague, and the chatbot produces the answer immediately.”

What separates an AI trading tool that helps from one that does not?

The data underneath it.

A conversational interface is straightforward to build. What determines whether the answer is worth acting on is what the interface is connected to.

“What matters is the data underneath it. A tool reading an overnight file will give you a clean, confident answer about yesterday. On a quiet day the difference may be immaterial. In a volatile session it is not.”

The work sits below the surface, and it is the part the market has been slowest to solve.

“Building the interface is no longer the difficult part of this. Having live, accurate data behind it is where the work actually sits, and it is the part most firms have not resolved. Our clients are not buying a conversational interface. They are relying on what it is connected to, and on the speed at which it captures their positions and aggregates that data.”

What does an AI chatbot change on a trading desk day to day?

Ask, instruct, receive.

The practical change is in how a desk reaches its own numbers, and what becomes possible once that friction is gone.

“Tommy gives you an interface in which you can speak directly to your data. You do not need to run calculations or go looking for numbers. You can simply ask, instruct and receive. That creates efficiencies, and it may also lead to new ideas and outputs.”

Where does AI go next in the Loqsea platform?

Beyond a single interface.

Tommy was a starting point. The direction of travel is towards clients reaching Loqsea data from wherever they already work.

“Tommy was the first step rather than the end destination. On the roadmap is the ability for our clients to reach their Loqsea data from the AI tools they already work in, Claude and others, rather than only within our own platform. That work is in development rather than live, and I would rather describe the direction of travel than commit to a date.”

The reasoning behind it is the same reasoning that produced Tommy.

“The principle behind it is the same one that produced Tommy. We capture and organise the data on behalf of our clients, and allow them to use the latest AI tools to analyse and compute that data into meaningful outputs.”

What is the latest at Loqsea and where is the focus?

Metals growth on a cross-commodity platform.

Development pace has been the defining feature of the past year. Metals has been the strongest growth area, supported by Loqsea’s LME ISV status, while the platform itself has always covered the full commodity derivatives range.

“The pace of development has been the defining feature of the last year, and that is deliberate. Metals has been our strongest growth area, helped by the LME work, and it is where a lot of attention has gone. Energy remains a significant part of what we do and the platform has always been cross-commodity.”

Attention is also shifting towards markets where trading volumes have outpaced the infrastructure supporting them.

“Beyond product, we are spending more time in markets where volumes are rising faster than the infrastructure supporting them. Those desks tend to carry less legacy technology, so when they decide something needs to change, they can act on it.”

Reza’s message to commodity trading firms

First mover advantage is real.

“First mover advantage is real. You want a forward-looking risk management platform that works with you, provides you with the latest technology, and equips you to make the most of your data. We are replacing legacy CTRM systems, and we are also paving the way to a new approach to risk management in the age of AI.”

Ready to see what an AI chatbot for commodity trading analysis looks like running on live platform data?

Schedule a demo and we will walk you through it in the context of how your desk actually runs.

Book a demo