Catalyst Raises $30M Seed to Build AI Trading Agents for Retail

Catalyst, a one-year-old startup building AI Trading Agents for everyday investors, announced on October 8 that it has closed a $30 million seed round led by Sequoia Capital, according to Fortune. The round, one of the largest seed financings of the year for a retail-focused AI startup, drew participation from Jump Trading, PeakXV, Lux Capital, AntiFund, Coinbase, and Premji Invest.

Founded in 2025 by CEO Justin Zheng, 25, and Dylan Iskandar, 21, Catalyst is betting that the next breakthrough in AI Trading Agents is not another hedge-fund automation tool but a consumer product that lets ordinary people describe what they want in plain language and have agents work out the execution. According to the company, the agent layer converts natural-language investment intents into full trading strategies, handling asset selection, cost optimization, and trade execution — with a human still required to confirm before any order is placed.

The announcement, reported by Pulse2, follows an unusually strong pilot. Catalyst says its early testing generated hundreds of millions of dollars in trading volume over the course of just a few weeks, and the company is now beginning to release users from its waitlist.

What Catalyst's AI Trading Agents actually do

Retail investors have had easy access to markets for years through web and mobile brokerages, but sophisticated strategies remain hard to implement. Translating an investment idea into an actual trade still requires navigating market infrastructure, venue selection, order types, and backtesting — a gap Catalyst's AI Trading Agents are designed to close. Users describe a thesis in plain language; the agents gather and analyze market signals, identify execution paths across platforms and venues, and help backtest and deploy the resulting strategy.

The company also incorporates limits and safety controls intended to give users greater visibility into how strategies are constructed and executed. According to the firm's own announcements, the platform combines AI Trading Agents with trading systems and financial infrastructure so users can evaluate opportunities, understand available options, and judge how a strategy could actually be executed.

The human-confirmation requirement is a deliberate design choice at this stage. The agents propose and prepare; the user approves. Iskandar has described the long-term vision in simple terms, telling Fortune that users will simply state an intent and the agents will handle everything end to end — but the current product still keeps a person in the loop, a constraint that distinguishes Catalyst from fully autonomous trading systems.

Sequoia's bet and the retail AI Trading Agents wave

The round lands in the middle of a crowded week for agent funding: Manus raised more than $500 million after its Meta acquisition collapsed, and Rein Security landed a $25 million Series A to secure the very class of autonomous agents Catalyst is selling to consumers.

The founders' backgrounds add texture to the story. According to AI Weekly's rundown of the raise, Zheng previously worked at Sam Altman's Worldcoin, while Iskandar published a cybersecurity paper with the U.S. Department of Defense while still in high school. The founding team's youth — a 25-year-old and a 21-year-old — recalls an earlier era of consumer fintech, when startups built by founders barely out of school rewrote how retail banking and brokerage worked.

Zheng has pushed back on the most obvious critique of retail AI Trading Agents — that they will function as a gamified casino — telling Fortune the company is building extensive user-education tools alongside the trading product. Iskandar has framed the mission around access, telling the magazine that a farmer should not need a hedge fund to be able to hedge.

The open questions

Skeptics note that several key details remain undisclosed. According to Wilson Sonsini's announcement of the transaction, which advised Catalyst on the round, the firm plans to extend its AI Trading Agents to a broader retail audience — but the company has not disclosed which brokerages handle execution, how it plans to make money, or how many pilot users produced the headline trading-volume figures.

There is also a genuinely hard product problem underneath the pitch. Natural-language instructions are far more ambiguous than traditional order entry: a user who says they want to buy a little bitcoin when it breaks a round number leaves the agent to resolve how much a little means, what counts as a breakout confirmation, and how stop-loss conditions should be set. That ambiguity matters beyond trading — a recent agent governance crisis has shown that a third of firms have acted on flawed AI output, which is why Catalyst's insistence on human confirmation is structurally important. The company says its AI Trading Agents resolve these ambiguities, but the details of that resolution — and which variables users can adjust during the confirmation step — have not been disclosed.

That ambiguity is precisely why the current human-in-the-loop design matters. As long as a person reviews and approves each strategy before execution, the agent functions as a sophisticated assistant rather than an autonomous trader. Whether Catalyst can safely move toward its stated end-to-end vision — while regulators, brokerages, and users all watch closely — will be the story to track as the waitlist opens. For now, the $30 million says that some of the smartest money in venture capital believes retail AI Trading Agents are the next consumer fintech frontier.