ExeQution Analytics Launches AI For Trading: Eolas

ExeQution Analytics, the specialist trading analytics company, has launched a hallucination-resistant AI assistant for trading - Eolas.

Eolas is an agentic tool designed specifically for trading, quant and IT teams in financial institutions to extract more value from their data and improve trading decisions.

Critically, Eolas has been engineered to be hallucination-resistant, ensuring its responses are drawn from a firm’s market or proprietary data, with no margin of error to fabricate responses. CEO Cat Turley said Eolas represented a step-change in institutions’ ability to monetise data investments.

“Eolas essentially removes all the bottlenecks firms grapple with around data,” she said. “It allows anyone involved in the trading function to speak directly to the data source, in natural language – and for the data to speak back.”

Eolas can be used to power best execution by allowing traders to interrogate realtime and historical data without going to quant or IT teams to pull reports, meaning intraday opportunities can be captured on the fly and issues mitigated before they drag on performance. For sales teams, Eolas can prepare client reports and insights in minutes that today take hours or days, improving responsiveness and service levels. For Heads of Trading, Eolas provides both the helicopter view needed to manage the floor as well as the ability to dive deeply into issues the moment they arise.

This ability for trading functions to use data without going to quants or IT relieves these teams of the burden of constant ad-hoc requests and regular reporting. This allows these valuable resources to focus on higher-value work, supported by the ability to interact with data far more quickly and easily themselves. Eolas also supports compliance and risk by allowing direct access to the data that surfaces risk or confirms adherence to rules and regulations.

Eolas is prevented from hallucinating by converting natural language queries into API calls. This limits it to performing approved functions on approved data, and ensures auditable execution paths, but still at the speed users expect from experience with other agentic agents. The API integration allows existing entitlement systems to be used, ensuring data is secure and only visible to permissioned users.

“The largest data set and most powerful analytics engine in the world is of limited use if only a few people can access it,” Ms Turley said. “Eolas is the next generation of trading intelligence, a true democratisation of data that can match the speed of insight to the speed of markets for the very first time.”

ExeQution Analytics has also released a paper, The role and future of AI on the trading floor <link>, discussing the role of agentic AI in trading.

Eolas architecture is based on the Model Context Protocol, Anthropic’s open-source integration protocol. It connects the client’s AI interface – typically via Claude, Gemini or CoPilot – to a KDB MCP server and then to ExeQution Analytic’s powerful analytics framework, built in q on KDB. The framework, and Eolas, are delivered as a bespoke service for each client.

ExeQution Analytics was launched in Australia in 2021, and has grown to service a global client base including hedge funds, sovereign wealth funds, propriety trading firms, asset managers and brokers, across all exchange-traded asset classes, with FX being added this year.