When Jacob Chanyeol Choi and his co-founders left MIT's orbit in 2022, they carried a thesis that now feels almost inevitable: institutional investors, drowning in data, were about to need something closer to a thinking partner than a terminal.
Two years and $6.6 million in seed capital later, LinqAlpha—the New York startup building what it calls an "Alpha Intelligence Layer" for buy-side firms—closed a $22 million Series A on July 2, 2026. AVP, the venture arm formerly known as AXA Venture Partners, anchored the round alongside Atinum Investment and GFT Ventures. The syndicate reads like a roll call of Asian capital markets: SBI Investment, Z Venture Capital, Samsung Securities, East Ventures, SV Investment, and the venture divisions of Mirae Asset joined in, along with Betatron Venture Group and NuVentures in India.
The timing is no accident. LinqAlpha is sprinting to scale just as the AI-for-finance category has lurched from experiment to arms race. AlphaSense, the category's older sibling, commanded a $7.5 billion valuation barely a month before LinqAlpha's announcement—a $350 million round in early June that underscored how quickly institutional appetites have shifted. FactSet pushed AI-powered document search to more than 85,000 users on March 26. Bloomberg and S&P Capital IQ Pro have spent the past two years retrofitting legacy systems with generative models.
Yet the founders insist they're building something structurally different—a multi-agent architecture that doesn't just accelerate research but surfaces market-moving signals before they're priced in. Whether that claim holds depends on execution, of course. But the client list suggests at least some traction: more than 70 institutions across three continents, collectively managing over $5 trillion in assets, including Causeway Capital Management and Schonfeld Strategic Advisors.
From Goldman to Multi-Agent Systems
The team carries a pedigree that mixes Wall Street with computational rigor. Co-founder Hojun Choi came out of Goldman Sachs' investment banking division. Subeen Pang, another co-founder, holds a PhD in computational science from MIT. Jacob Chanyeol Choi has spent years organizing AI-for-finance events with quantitative heavyweights like AQR. Jin Kim earned a Forbes 30 Under 30 nod in the AI category.
They launched as simply "Linq" before rebranding to LinqAlpha in 2024—a shift that coincided with the company's pivot toward what it describes as an intelligence layer rather than just another research tool. The platform pulls in filings, alternative data feeds, and internal proprietary research, then stitches them into a single interface with source-linked outputs. The pitch, essentially: let machines chase the signals while analysts chase the insights.
"The first wave of AI in finance made analysts faster," Hojun Choi said in the announcement. "The next wave changes what they can know."
That framing resonated with Manish Agarwal, General Partner at AVP, who positioned the investment as a bet on differentiation over speed. "LinqAlpha is addressing a larger opportunity: building systems that help institutional investors discover differentiated insights," he said. Translation: this isn't about replacing Bloomberg terminals with chatbots. It's about finding asymmetries that legacy infrastructure can't.
The Validation Engine
On June 25, LinqAlpha collected the "Best AI Solution" trophy at the 2026 Hedge Fund Services Awards, organized by With Intelligence—a recognition that carries weight in a category where credibility moves slower than technology. Earlier this year, the company contributed to AWS's machine learning blog and joined Microsoft's Majung Program, which focuses on building secure, Azure-native financial AI infrastructure.
The company now operates hubs in Korea, the United States (including Cambridge, MA), Hong Kong, and Singapore. Its website claims access to primary sources across 139 countries in real time—a figure that sounds sweeping but is difficult to verify independently. The platform already offers API endpoints and what it calls "MCP" interfaces for weaving financial data into third-party AI applications, a nod toward becoming infrastructure rather than just software.
The Korean business press took particular notice of the round, highlighting the participation of Samsung Securities, NH Investment & Securities, Mirae Asset, Shinhan Venture Investment, and Hana Ventures—a cluster of local players that telegraphs distribution ambitions in Asian markets. Some Korean outlets reported the raise in won terms at roughly 34 billion KRW, a figure slightly higher than the $22 million USD headline, though LinqAlpha's official announcement treats the dollar figure as authoritative.
Building Against Decades of Incumbency
Perhaps the most revealing detail is what LinqAlpha hasn't disclosed: a specific roadmap for deploying the $22 million. The company's public statements emphasize expanding its "Alpha Intelligence Layer" and developer tools, but the absence of concrete milestones leaves room for interpretation. That's not necessarily a red flag—many startups at this stage guard their playbook closely. But it does underscore the challenge: LinqAlpha is building in a category where incumbents have decades of infrastructure, client relationships, and switching costs baked in.
The bet, stripped to its core, is that a multi-agent architecture purpose-built for today's AI models can carve out meaningful space against legacy terminals. That proposition hinges on whether the platform can keep surfacing what investors don't yet know—and whether it can do so faster, more reliably, and with enough edge to justify the disruption of ripping out entrenched workflows.
With $28.6 million raised to date and a client base managing trillions, LinqAlpha has bought itself runway. Whether it's enough to outrun the category's established giants remains the open question. The market is moving quickly. So, evidently, is LinqAlpha.
