Catalyst Raises $30M Seed for Retail AI Trading Agents

El Niño sits on Catalyst's homepage as a trading prompt. A weather pattern becomes the starting point for software that turns a person's market view into a strategy, watches for conditions and can execute actions inside the boundaries that person has chosen.

Catalyst has raised a $30M Seed round led by Sequoia Capital, announced on October 8, 2026. Co-founders Justin Zheng and Dylan Iskandar are building AI trading agents for retail investors, with Jump Trading, PeakXV, Lux Capital, AntiFund, Coinbase and Premji Invest also participating. The financing supports a company aiming to make more of the work between an investment idea and an executable strategy accessible through ordinary language.

The decision hiding inside a trading prompt

A person with a market opinion still has decisions to make about how to express it. The instrument, timing, budget and conditions for changing course all influence what the resulting trade will do. Catalyst brings those choices into a visual workflow, placing the construction of a strategy close to the conversation that started it.

That makes explanation part of the product. A useful interface has to help a customer understand the exposure created by the instruction, including what the software will keep watching after setup. Easier language can make a complicated decision feel finished sooner than it actually is. The workflow gives the customer somewhere to inspect how their intention has been translated.

Sequoia's investment announcement describes gathering signals, finding routes across platforms and venues, and helping users backtest and deploy strategies with limits. Those are the investor's descriptions of Catalyst's approach. They explain why the company is working beyond a conversational research screen, into the sequence that connects an idea with financial action.

The customer relationship changes along that sequence. Research can inform a decision the person makes elsewhere; execution brings the product into the decision's consequences. Catalyst's commercial ambition depends on making that handoff understandable to people whose knowledge of an asset may be much stronger than their knowledge of trading systems.

What the Seed financing establishes

The transaction's label is worth getting right. Both Sequoia and the law firm that advised Catalyst identify the financing as Seed. Wilson Sonsini confirms the $30M amount, October 8 announcement and participating investors. The available primary disclosures provide a clear funding event without a valuation or an allocation of each investor's commitment.

Sequoia's involvement places a venture investor alongside firms with financial-market and technology exposure in the syndicate. Their participation gives Catalyst resources for its broader retail ambition, while leaving the product team to decide how that ambition becomes usable software. The announcement does not specify a timetable for general availability or spending by engineering, distribution and customer education.

The founders remain central to that job. Sequoia names Justin Zheng and Dylan Iskandar as co-founders and discusses the need to bridge customers' understanding of financial instruments. Its account is an investor's explanation of the opportunity, so the team's capabilities should be assessed through the experience users can actually inspect. No CEO or CTO title allocation is needed to explain the responsibility the founders are taking on.

For an operator reading the financing, the useful question concerns adoption. A larger audience will bring a wider range of intentions, experience levels and assumptions into the same interface. Each strategy has to remain legible as that audience grows, including to the person who asked for it in the first place.

A strategy carries continuing permission

Catalyst's official FAQ describes spot tokens, perpetuals, tokenized stocks and prediction markets as supported markets. It says strategies can combine real-time signals and trading actions in a workflow. These are company-described capabilities, and the range itself raises a design question: how clearly can someone see what a single instruction has set in motion across different markets?

The company says users choose budgets, strategies and execution conditions, and can pause or modify those strategies. It describes the service as non-custodial. The important operating distinction is that an authorized workflow can continue acting automatically within its parameters. A customer should not be pictured approving every individual trade merely because the initial strategy began with a conversation.

Catalyst's terms of service identify Catos Labs, Inc. as the service company and describe agents acting under users' instructions, permissions and parameters. The terms also state that the company is not registered as an investment adviser with the SEC or as a commodity trading advisor with the CFTC, and that regulatory authorities have not approved the service. Those disclosures belong alongside any description of automation; they prevent a convenient interface from being mistaken for a regulated advisory relationship.

The operating consequence follows directly from continuing permission. A market view can change while the workflow remains configured around the earlier view. Pause, modification and clear visibility into the strategy therefore occupy the same customer experience as setup and execution, especially when the person is no longer watching the screen.

Retail reach makes comprehension a distribution problem

Catalyst's current website offers a waitlist, while its FAQ says the research preview is free. Wilson Sonsini describes plans to extend the product to a broader retail audience. Those observations establish the direction of travel without assigning a commercial launch date or assuming that the audience already reflects the full ambition.

For the founders, broader reach means reaching people with different ways of expressing the same idea. An experienced trader may specify limits immediately; someone entering from a news story may describe only the expected outcome. The interface has to keep enough of the strategy visible for both people to recognize what they are authorizing. That is a customer-education challenge embedded in distribution rather than reserved for a separate help page.

The financing gives Catalyst a larger budget around this work, and the product provides a concrete place to observe it. A plain-language instruction becomes a workflow with signals, conditions and financial actions. A retail user arriving from a weather forecast will bring their own idea of what the forecast means. Justin Zheng and Dylan Iskandar are asking that person to remain an informed participant as an agent converts the idea into instructions the market can execute.