The San Francisco startup connects to the brokerage accounts you already have, reads what you own, and researches the market while you're doing something else. Its bet: retail investors don't need another buy button - they need a second opinion.
More than 160 million Americans own stocks. According to Astor's own reading of the market, more than half of them don't feel confident about the decisions they make with that money. That gap - between owning a portfolio and actually understanding it - is the whole reason the company exists.
Astor is an AI investment advisor. Not a chatbot bolted onto a trading app, and not a robo-advisor that quietly rebalances in the background. It connects to the brokerage accounts a person already uses, evaluates their holdings across performance, risk and diversification, watches the market around the clock, and then explains what it finds in plain language through chat and voice. The company came out of Y Combinator's Summer 2025 batch and, in April 2026, announced a $5 million seed round led by Monashees.
The tagline the founders landed on is refreshingly literal: "Finally, your portfolio answers back."
Bruno Koba, Astor's co-founder and CEO, describes the moment the idea clicked in unglamorous terms. "We looked around and everyone we knew was investing on their own," he has said, "with many treating their brokerage accounts like a casino." That is the problem in one sentence. The average person can open an account in minutes and buy almost anything. What they can't easily get is the thing professionals take for granted: someone reading filings, tracking catalysts, and telling them what actually matters for the specific stocks they hold.
The tools most retail investors use to fill that gap have barely changed in fifteen years. A ticker, a chart, a red-or-green number, a buy button. Astor's argument is that the interface itself is out of date - that the natural way to interrogate a portfolio in 2026 is to ask it questions and get answers back, not to squint at a candlestick chart and guess.
The mechanics are straightforward, which is part of the appeal. A user links the brokerage account they already have - custody of the money never leaves the broker. Astor reads the holdings, then goes to work: monitoring the market 24/7, filtering the firehose of news down to what's relevant to that particular portfolio, and flagging developments in real time.
Link an existing brokerage account. Custody stays with your broker.
Astor reads holdings across performance, risk and diversification.
Markets tracked around the clock; alerts filtered to your portfolio.
Tailored research and recommendations, by chat or voice.
Underneath the conversation sit a handful of specific jobs. There's catalyst detection, which surfaces unusual insider activity and emerging opportunities before mainstream coverage catches up. There are tailored research briefs, which compile data, filings and expert commentary into a single decision-ready report built around what a person owns. And there's scenario testing - a "what if" mode that lets an investor stress-test a hunch against supporting market data before putting real money behind it.
Put those pieces together and the product starts to look less like an app and more like a research desk that happens to fit in a pocket. The point of institutional research teams isn't that they read faster than everyone else. It's that they read constantly, and they know which few facts out of ten thousand actually move a position. Astor is trying to compress that judgment into software - the filtering, the framing, the "here's why this matters to you specifically" - and then hand it to someone who has a day job and forty minutes on a Sunday to think about their money.
The chat-and-voice interface is doing more work than it looks. Asking a question out loud, or typing it the way you'd text a friend who happens to know markets, removes the biggest barrier for most people: not knowing what to ask, or where to look, in the first place. You don't need to know the name of a filing to say "did anything weird happen with the companies I own this week." That's the part Astor is betting is genuinely new.
Here is the detail that separates Astor from the crowd of AI finance apps: it is a registered investment advisor with the SEC, operating under fiduciary duty. In plain terms, it is legally obligated to act in its clients' best interests. Plenty of products will happily give you "information" and disclaim any responsibility for what you do with it. Astor built the regulatory weight in from the start, and made it a selling point rather than a footnote.
That choice shapes everything else. A robo-advisor rebalances silently and rarely explains itself. A stock-picking newsletter has opinions but no obligation to you. Astor sits in a different spot - a conversational analyst that shows its reasoning, that you can push back on, and that carries a legal duty to be on your side.
The business model follows from the structure. Astor doesn't take custody of anyone's money; it advises on accounts that stay put at the broker where they already live. That keeps the relationship clean and the switching cost low - a user can connect Astor without moving a single share. As a registered advisor, the natural way to charge is a subscription or advisory fee rather than the payment-for-order-flow model that powers much of commission-free trading. It's a quieter way to make money, and arguably a better-aligned one: Astor earns when the advice is worth paying for, not when you trade more.
Illustrative positioning, not a ranking.
Astor's two founders are Brazilian fintech builders, both originally from Sao Paulo. Bruno Koba, the CEO, was a data scientist at Nubank - the Latin American digital bank - where he built machine learning models that touched millions of customers. He later crossed the table to become a fintech investor at Monashees, the firm that would eventually lead Astor's seed round. He finished a Stanford MBA before going all in on the company.
His co-founder and CTO, Daniel Tulha, is a software engineer whose resume reads like a tour of modern financial infrastructure: Stripe, Robinhood, and Amazon. That combination - a data scientist who has also allocated capital, paired with an engineer who has built the plumbing behind payments and trading - is a specific kind of pedigree for a specific kind of product.
Downloads are easy to inflate. Connected accounts are not. Within months of launching, Astor reported thousands of users and more than $200 million in linked brokerage assets. When someone connects their actual portfolio to a new app, they're making a small statement of trust - and that trust is the real currency for a company asking to be your financial second opinion.
The seed round reads like a vote from people who know the space. Monashees led, with Y Combinator, Goodwater Capital, Gilgamesh Ventures, 468 Capital, Valutia and Sunshine Lake alongside, plus individual executives from Stripe and OpenAI. The stated plan for the money is unremarkable in the best way: grow the product, engineering and growth teams, and expand what the service can do.
Astor is standing on an interesting seam. On one side is the giant, sleepy world of retail brokerage, where the interface has barely evolved. On the other is the fast-moving world of AI agents, where software can now read, reason and summarize at a level that was science fiction a few years ago. The company's whole thesis is that these two worlds are about to collide, and that the winner will be whoever makes institutional-grade research feel as simple as sending a text.
It won't have the space to itself. Robo-advisors like Betterment and Wealthfront own the "set it and forget it" crowd. Robinhood and the big brokerages own the accounts. Research platforms like Seeking Alpha and Morningstar own the deep-dive reader. Astor's wager is that none of those quite scratch the itch of an active investor who wants a knowledgeable, always-available second opinion tied to their own holdings - and who would rather ask a question than read a 40-page report.
Whether that becomes a habit millions of people build is the open question. For now, Astor has a clear pitch, a technical team that has shipped financial products before, real money in connected accounts, and a name for what it's trying to be. Not a casino. Someone in your corner.