Kalshi has developed an AI agent to manage internal processes including the review of prediction market contract language, co-founder Luana Lopes Lara said in an interview with Bloomberg.

The tool, known internally as Harrison, helps the company handle millions of daily wagers on outcomes of events like elections, sports games and award ceremonies.

Bets often depend on the specifics of how Kalshi’s contracts are written, including language and evidence sources. The AI agent performs daily tasks like aggregating top news, analyzing competitor offerings and making recommendations on what the exchange should list next or where it should focus rewards for users adding liquidity.

Staff on Kalshi’s markets team are the biggest users of the technology outside of engineering among the company’s 150-person workforce, Lopes Lara said. The agent is built on top of Anthropic’s Claude model.

Kalshi now has more than 500 templates for possible markets that have been reviewed by its team, Lopes Lara said. Each template goes through review for how it can be generalized to fit more events, how it can be stress-tested and whether it meets user requirements.

Determining an outcome is a three-step process in most cases. One person on the markets team inputs an outcome into the system while a second person independently adds their own decision. Kalshi’s AI verifies whether the answers match while also checking against its own suggested response.