Bruno Koba spent years watching wealth managers ignore an entire class of earners. Not because they weren't making good money—many were pulling in six figures—but because they didn't yet have enough of it stashed away to clear the minimums that white-glove advisory firms demand.
Now Koba thinks he's found a way in. His startup, Astor, charges $15 a month to deliver what it bills as SEC-registered investment advice, powered by AI and aimed squarely at the cohort known in wealth management circles as HENRYs: high earners, not rich yet. On April 23, the San Francisco company announced it had closed a $5 million seed round led by Monashees, with participation from Y Combinator, Goodwater Capital, and a handful of other investors, including unnamed executives from Stripe and OpenAI.
The pitch is deceptively simple. Link your brokerage account, let Astor's algorithms chew through your portfolio, and receive personalized recommendations—delivered via text or voice through a conversational interface. You still execute the trades yourself; Astor operates as a non-discretionary advisor, which means it suggests but doesn't pull the trigger. The company uses Anthropic's language models in what Koba describes as a multi-agent setup, though the technical architecture remains largely a black box to outsiders.
What's less simple: whether anyone will trust an algorithm with their retirement savings, even at $15 a pop.
A Crowded Field, Moving Fast
Astor is hardly alone in chasing this opportunity. Robinhood, the brokerage that made stock trading frictionless for millions, has been publicly communicating about its AI assistant Cortex as of December 2025. Perplexity, the search startup, rolled out Plaid-powered portfolio integrations this past spring. Origin, which serves both enterprise clients and retail customers, launched its own SEC-regulated AI advisor late last year.
The difference, according to Astor, is fiduciary status. The company operates under the name Gaus, Inc., registered with the SEC (CRD #338424), and holds itself out as a fiduciary under the Investment Advisers Act—a designation that legally obligates it to act in clients' best interests. Koba himself holds a Series 65 license, issued in January 2026. It's a credential that requires passing a rigorous exam and, more importantly, signals that Astor isn't just a chatbot with a finance wrapper.
Still, skepticism abounds. Fiduciary duty is one thing; execution is another. Can a startup with five employees—per Y Combinator's company directory—really deliver the kind of nuanced, tax-aware, life-stage-appropriate advice that justifies even a modest subscription? And what happens when the models hallucinate, or miss a key risk factor buried in an earnings report?
The Numbers Game
Since launching its iOS app in February, Astor claims to have attracted thousands of users and more than $200 million in connected accounts, according to an April press release. Fortune pegged the user count at around 4,000 as of the funding announcement. A day later, the company's website displayed a different set of figures: 5,000+ users and $300 million+ in assets under advisement. Fast-moving metrics for an early-stage product, or creative rounding? Perhaps both.
Pricing runs $15 per month for the standard tier. A higher tier, reportedly around $40 monthly, offers unlimited access—though what exactly that unlocks remains vague. For context, traditional robo-advisors like Betterment and Wealthfront typically charge a percentage of assets under management, usually around 0.25%. A $100,000 portfolio would cost $250 a year on that model, or $180 annually with Astor. The math works, assuming the advice is comparable.
Unlikely Backers, Familiar Faces
The seed round's composition is revealing. Monashees, the Brazil-based venture firm, led the deal—notable in part because Koba previously worked there as a fintech investor before pursuing an MBA at Stanford, which he completed in 2025. It's the kind of investor-founder relationship that raises eyebrows in some circles, though it's hardly unusual in venture capital.
Joining Monashees are 468 Capital, Valutia, Sunshine Lake, and the usual YC contingent. The presence of operators from OpenAI and Stripe on the cap table—albeit unnamed—hints at the technical and regulatory complexity involved. Shipping AI-driven financial products at scale isn't just an engineering challenge; it's a compliance minefield.
Koba's co-founder, Daniel Tulha Hochstetler, brings his own pedigree: stints building software at both Stripe and Robinhood, two companies that have redefined fintech infrastructure over the past decade. The duo went through Y Combinator's Summer 2025 batch, which has become something of a finishing school for founders aiming to crack highly regulated industries.
The Trust Problem
What Astor is really selling isn't algorithms or even advice. It's trust, packaged for a generation that grew up with apps but inherited their parents' anxiety about money. The HENRY demographic—often defined as individuals earning between $100,000 and $250,000 annually, with net worths still shy of the million-dollar mark—sits in an awkward middle ground. Too affluent for the generic guidance offered by free tools, too broke (relatively speaking) to justify paying a wealth manager's fees.
The question is whether a $15 subscription can bridge that gap. Astor's early traction suggests there's genuine demand, but demand and defensibility are different beasts. Regulation provides some moat—you can't just spin up an SEC-registered investment advisor overnight—but every major brokerage and fintech platform is already moving in this direction. Robinhood has distribution. Perplexity has brand momentum. Origin has enterprise relationships.
Astor has speed, perhaps, and a focused thesis. Whether that's enough to build a durable business in a category where accuracy, compliance, and customer trust all matter equally is the bet Monashees and its co-investors are making. The company plans to use the fresh capital to expand its engineering team, build out additional service lines, and invest in growth initiatives—venture-speak for "hire, ship, and acquire users faster than the competition."
For now, the market seems willing to give Koba and his team the benefit of the doubt. How long that goodwill lasts may depend less on the sophistication of Astor's AI than on whether it can avoid the kind of high-profile blunder that ends careers in regulated finance. One bad recommendation, one viral complaint, one regulatory slap—and the whole experiment could unravel.
Then again, at $15 a month, maybe the risk tolerance is higher than anyone realizes.
