Builds AI trading agents that turn natural-language intent into executable, human-approved trading strategies for everyday investors.

Retail investors will state goals in plain language and let AI agents design, cost-optimize and execute the whole trading strategy — with a human approving each trade.

$30 million seed led by Sequoia.

Justin Zheng, 25, got into Ethereum in middle school, built a Village Global-backed biometrics startup in high school and joined Sam Altman's Worldcoin in 2019; Dylan Iskandar, 21, published cybersecurity papers for the US Department of Defense in high school. They met over a 2019 poker game at a Tyler Cowen conference, stayed in touch, and circled one question — how to make an increasingly complex financial world navigable for everyone — before cofounding Catalyst in 2025.

Sequoia led the $30M seed reported October 8, 2026, joined by Jump Trading, PeakXV, Lux Capital, AntiFund, Coinbase and Premji Invest. The agents morph natural language into a trading strategy and then execute it, with a human in the loop accepting each trade. A pilot focused on power traders generated hundreds of millions in trading volume over a couple of weeks, by the company's account; waitlist users are being let in now, with wide availability expected over time.

Assets and strategies have outpaced human comprehension; an intent-to-execution agent collapses that gap for people without a hedge fund.

Power traders were a sharp wedge: they generate volume and stress the execution engine before retail trust exists.

Human-in-the-loop acceptance answers the safety objection directly — the agent proposes, the person disposes.

The investor mix pairs trading expertise (Jump Trading) with distribution and mainstream legitimacy (Sequoia, Coinbase, Premji Invest).

When a domain outgrows any individual's comprehension, the winning product is not more data but an agent that turns intent into execution — with a human kept in the loop and the casino kept out.

As of October 8, 2026, Catalyst was letting users off its waitlist with wide availability expected over time; the pilot volume figures are self-reported and the product has not been independently audited. Fortune noted the platform launches into a regulatory and cultural moment where prediction markets and crypto are normalizing — and misuse is an open question.

FOLLOW THE EVIDENCE

The sources

  1. Sequoia-backed Catalyst raises $30 million to build AI trading agents for everyday investors fortune.com