← /editions
#121

Agents Start Moving Money: Catalyst's $30M Seed and Instinct's $10B Bet

// Sequoia leads a $30M seed for Catalyst, whose retail trading agent drove hundreds of millions in volume in a two-week pilot, following Instinct's $1B raise at a $10B valuation. Capital is flowing to agents that execute, not just advise.

$ today_in_10_seconds

Agents are moving from answering questions to moving money. Catalyst says its retail trading agent drove hundreds of millions in volume in a two-week pilot, and Sequoia led its $30M seed. Sequoia also co-led Instinct's $1B round at a $10B valuation late last month. The smart money is pricing in agents that execute, not just advise.

$ lead_story

Catalyst raises $30M seed for an AI agent that trades for retail investors

Catalyst, a startup building an AI trading agent for retail investors, raised a $30M seed led by Sequoia. It says a pilot generated hundreds of millions of dollars in trading volume over just two weeks. That's an agent with real autonomy over real capital, which raises the stakes on everything builders usually hand-wave: guardrails, user confirmation flows, auditability, and what happens when the agent is confidently wrong. Keep in mind that volume isn't returns. The number shows people will let an agent act on their money, not that the agent makes them money.

// why_it_matters: Financial execution is the hardest trust test for agents, and users are apparently already passing it.

read more →

$ also_worth_attention

  • Instinct raises $1B Series C at a $10B valuation, rolls out a concierge service

    AI agent company Instinct raised $1B from Sequoia, Benchmark, and Coatue on September 28, and pointed to new products including a concierge service. A concierge is the consumer-friendly name for a general-purpose agent that takes multi-step actions on your behalf. At $10B, investors are betting Instinct can own that delegation layer before the platforms do.

    link →

$ bigger_picture

The trend behind the headlines

Sequoia shows up on both cap tables, and the pattern is clear: capital is flowing to agents that complete transactions, not chat interfaces that suggest them. One bet is narrow and high-stakes (trades), the other broad and everyday (errands). Both depend on the same unsolved problem, which is getting users to grant an agent standing permission to act without approving every step. Whoever builds the best permissioning, rollback, and accountability layer will matter as much as whoever has the best model. Expect regulators to take a close look at agents placing retail trades.

$ tool_of_the_day

Alpaca MCP Server

If Catalyst has you curious, this lets an MCP-compatible agent query market data and place orders through Alpaca's brokerage API. It also supports paper trading, so you can stress-test your agent's guardrails before it touches a real dollar.

explore Alpaca MCP Server →

$ subscribe --free

Get the next edition in your inbox

// DAIly Agent covers agentic AI three times a week

// free, no spam, exit anytime