Goodfin is an example of a platform built by a small team that even a few years ago would have been impractical or unthinkable: A (for now) direct-to-accredited-investor, private markets platform focusing on late-stage, pre-IPO and venture/growth private equity opportunities.
Its creators refer to it as an agentic wealth platform. That’s because it has been built atop an agentic framework from the start, and its most recently launched and first tool, Goodfin Go, is an AI agent-to-human model that enables the agent to perform complex research, legal document queries and portfolio monitoring (among other things) in real time.
Stay with me. On the future roadmap is a version for advisors.
“I built it because it is what I wanted at the time and could not find,” said Anna Joo Fee, Goodfin’s Harvard-educated founder and CEO. Fee was an attorney who had been working on private market deals, and several years ago she started thinking about what would become Goodfin.
Spelled out in the first few sentences of its homepage disclosure, the company makes clear that it is a technology company, not a broker/dealer or RIA.
The 10-person startup, which is backed by Y Combinator, launched in 2022 and emphasizes a social and community-driven approach and is targeting startup founders and professionals who want to own a piece of the future by investing in pre-IPO ventures; those interested in the space (SpaceX is available on the platform), robotics and AI industries, for example, said Fee.
A screenshot of Goodfin Go.
“We’re building everything from onboarding to market research, kind of like an analyst for private markets in your pocket,” she said referring to Goodfin Go, a system with more than 30 AI “specialist” agents that work together on everything from research, to analysis, as well as compliance-related tasks, and even executing transactions (though investor members still have to wire funds into their accounts, Fee said).
Being direct-to-consumer (albeit for accredited investor consumers), the tools have been designed from the start to personalize the experience, taking into account risk tolerance, what the system has learned over time about a user’s investing interests and habits, the sectors of interest and tracking deals the investor turned down. Fee said this is as much about customer service, in that the technology is learning over time what users like and delivering it to them without them having to manually do the research themselves. (Many advisors will likely see some obvious value in this for themselves in the future.)
I asked her to elaborate on the social and community aspect she had mentioned. She talked about the built-in collective intelligence the company’s engineers and developers had baked into the analysis process, which pulls in real-time sentiment from other Goodfin members (anonymized, non-personal data) who are actually investing, as well as those considering it, and their likelihood of investing (called conviction scores).
Then there are the compliance and execution-related tasks the agents can carry out, including verifying know-your-client and accreditation requirements and getting the required signatures. It still requires investor oversight to make and complete transactions, but it is not beyond the technology to automate this as well.
Fee reiterated something she told me during our first meeting, that Go was stress-tested by CFA-certified human advisors before it launched. And such advisors worked with Goodfin’s developers to roll out earlier platform features that are now part of agents within Go, including Deep Research Analyst, its AI Concierge and Investor Ticker.
Some advisors may recall the Goodfin name from its collaboration with NYU’s Stern School of Business on a study testing various Large Language Models to attain CFA Level III competence. (A couple achieved impressive scores, and the conclusion was that LLMs at present could “significantly augment the abilities of existing financial professionals.”)
“We started with private markets as a wedge,” she said, noting that so much still needed to be built to support that research and those transactions.
“But wealth is next,” she said of pivoting into supporting the wealth management side of financial services, though she could not at present provide a timeline.
As for competitors, it includes the wide and varied range of access investors currently have to the full spectrum of private markets, including private equity, private credit, real estate and infrastructure, namely through intermediaries, whether that is units within wirehouses and brokerages or platforms RIAs rely on for access to private and secondary markets.
Fee, when asked, did acknowledge that Arta Financial, the Singapore-based digital family office and private bank I wrote about last year, is a direct competitor.
“But we have different approaches,” Fee said, “We are focused on AI-native; we are focused on doing an average of $45,000 per investment, and 3 1/2 investments per user per year.”
While she declined to share the total number of users, she did say it is in the thousands, and the tempo and average deals per user per year had risen to 3.5 from 1.1 in 2025.
“We are very sticky,” she said, a term that those in Silicon Valley love to use in reference to customer retention and repeated use of their products or platforms.
And she was not short on examples supporting the case for continued demand and expansion in pre-IPO private equity access, from Robinhood’s launch of a special purpose vehicle for non-accredited investors in Europe (because U.S. regulators do not allow it) to Morgan Stanley’s acquisition of EquityZen (and there is also Schwab’s recent purchase of Forge Global).
For now, Goodfin remains by invitation or referral only and has two levels of membership: the premium entry level, which costs $750 annually (though that does not include deal fees), and the reserve membership for those allocating $100,000 or more annually across their private market investments.
The latter can vary in price, and Goodfin is not publishing it currently, but it does offer reduced or waived fees on select reserve-only deals and provides community access to other reserve members, as well as networking perks such as invite-only events and bespoke deal sourcing.
“Our goal is to build an entire suite of tools and products for our users,” Fee said.
