Sparta has launched Leonidas AI, a real-time decision-making engine designed for oil and commodity trading desks. 

The platform combines Sparta’s market data, live prices, arbitrage economics, news, fundamentals and seasonality to produce trade calls and explain the reasoning behind them.

The company said the tool addresses a growing challenge for trading desks: turning large volumes of market information into an actionable view quickly enough to capture trading opportunities. 

Leonidas continuously interprets and weighs market inputs, reducing the need for traders to manually connect curves, spreads, news and other signals.

The platform introduces dedicated AI agents for individual commodity markets, alongside calls, market briefings, research tools, market views and alerts. 

Traders can question curves, spreads, routes and positioning scenarios, while challenging individual elements of the analysis, including arbitrage economics, news weighting and seasonal assumptions.

Leonidas also monitors desks’ coverage and rolling market baselines, flagging abnormal moves and identifying potential drivers. 

Users can set thresholds for prices or spreads and receive alerts when they are breached, with supporting reasoning attached.

Sparta said the system personalises itself to each desk’s coverage, regions and thresholds over time. 

A tenant/landlord architecture keeps each organisation’s data and conversations isolated, preventing one desk’s information from influencing another’s analysis.