Catalyst, a San Francisco startup building artificial intelligence agents that convert conversational ideas into automated trading strategies, said Tuesday it has secured $30 million in seed funding led by Sequoia Capital. The round included participation from Jump Trading, Peak XV Partners, Lux Capital, AntiFund, Coinbase, and Premji Invest, according to Fortune, with Selini Capital and SF1 also participating according to the company's social media.
The company, which began releasing users from its waitlist following a pilot program, offers a platform where traders describe their market thesis in ordinary language. Its AI agents then gather signals, construct strategies, and execute trades across stocks, crypto assets, and prediction markets within boundaries set by users.
Catalyst is making its service free during what it calls a research preview period.
From Idea to Execution
The platform's core proposition is straightforward enough: users articulate a trading hypothesis the way they might explain it to a colleague, and the system handles the rest. According to Catalyst's website and Sequoia's announcement, the agents work across spot tokens, perpetual futures, tokenized equities, and prediction markets. Strategies can pull from market data, news feeds, social media activity, and sentiment analysis spanning multiple venues.
Under the hood, Catalyst operates with a non-custodial structure. Wallets are managed through Privy, and the system includes what the company describes as human approval gates, according to terms of service documentation. Users fund accounts via Apple Pay, credit cards, or various cryptocurrencies including USDC, SOL, ETH, and BNB. Withdrawals currently process in stablecoins, though the company says fiat withdrawal options are in development.
The terms make clear that Catalyst isn't registered with the Securities and Exchange Commission as an investment adviser or with the Commodity Futures Trading Commission as a commodity trading advisor.
The Builders
Co-founders Justin Zheng, 25, and Dylan Iskandar, 21, bring résumés that read like a Silicon Valley fever dream. Zheng launched a Village Global-backed biometrics startup while still in high school, then joined Sam Altman's Worldcoin project in 2019. He became the youngest fellow at IDEO CoLab on two separate occasions, according to Sequoia.
Iskandar published cybersecurity research for the U.S. Department of Defense before finishing high school, then arrived at Stanford as a Jane Street scholar. He left to build Catalyst rather than complete his degree.
The broader team includes alumni from Palantir, the first engineer at Scale AI, and technical leads previously at Citadel, Citadel Securities, Hudson River Trading, and Jump, Sequoia noted in its announcement.
Crowded Territory
Catalyst isn't alone in betting that conversational interfaces will reshape retail trading. Robinhood rolled out "Agentic Trading" and "Robinhood Chain" in mid-2026, while SoFi acquired Composer in June and rebranded it as an AI-powered investing tool that similarly builds strategies from natural language prompts. Earlier platforms like Capitalise.ai have offered code-free automation using plain language for years.
George Robson, the Sequoia partner backing the deal, framed Catalyst's agents as "a suit of armor as you navigate the financial world" in comments to Fortune. The firm's thesis centers on the idea that sophisticated market exposure should be accessible through everyday language rather than requiring technical expertise.
What Comes Next
"The idea is that people will simply state an intent, and we have our agents to figure out everything end to end," Iskandar told Fortune. His co-founder struck a cautious note about the platform's purpose: "We're not here to build a gambling app," Zheng said. "We want to be the ones that can safeguard your cash and help you reach your goals."
The company is methodically expanding access beyond its pilot group as it works through its waitlist. Broader availability will roll out over time, though Catalyst hasn't specified a timeline for general access.
