AlphaSense, a Goldman Sachs–backed AI research startup valued at $2.5B, gears up for IPO as it crosses $200M in annual recurring revenue

Hedge fund manager Cynthia Paul started using the research platform AlphaSense more than 10 years ago while she was working at George Soros’s family office.

She had enlisted all the analysts on her team to track down a data point—and one of them came back so quickly, showing her an answer on his computer screen, that she almost didn’t believe him. The analyst had found the figure via AlphaSense’s AI-powered search engine, which uses machine learning to pull from a dataset of public documents and SEC filings, transcripts, and research.

“I was impressed enough to try the product myself,” says Paul, who would soon ask AlphaSense if Soros Fund Management could invest in the firm, and would later back the company again when she launched her own hedge fund. Research startup AlphaSense now boasts more than 4,000 corporate customers, including 80% of the S&P 100, who typically pay between $10,000 or $20,000 annually per seat to follow their competitors, track regulatory approvals, or research M&A targets.

With the help of developments in generative AI, AlphaSense CEO Jack Kokko says he is seeing heightened interest in the platform—both from its users, and its swath of investors, ranging from Goldman Sachs, CapitalG, and Viking Global. AlphaSense hit $200 million in annual recurring revenue at the end of last year—doubling its ARR from summer 2022—and Kokko is now thinking about how to double that figure, while setting his sights on the right time to take the company public. “We’re suddenly a lot closer to that,” Kokko tells Fortune in an interview.

AlphaSense has increased its headcount by 20% since January 2023 and opened an office in Singapore. This month, AlphaSense hired a new chief marketing officer, Heather Zynczak—who was part of the executive team that took software company Pluralsight public—as it tries to lay the groundwork for an IPO.

“There’s still a lot to do, of course, when you actually decide to go and file,” Kokko says.

‘The poster child’

Twenty-five years ago—during the roaring dotcom boom—Kokko was working as an investment banking analyst for Morgan Stanley. He remembers the feeling of being a young analyst and walking into a client boardroom, his mind running over what data point he might have missed in his research—what overlooked item might end up “messing up that billion-dollar deal,” he says.

“That’s what stuck with me,” Kokko says, and what would ultimately lead him and his Wharton School classmate, Raj Neervannan, to start working on a search engine that could marry the burgeoning technology of artificial intelligence with research about companies.

The key to the AlphaSense platform, which Kokko and Neervannan started building in 2011, would be its dataset. AlphaSense’s models pull from SEC filings, equity analyst research, industry journals, patent filings, earnings call transcripts, and other public documents and sources. Its two acquisitions over the years—financial analysis company Sentieo as well as expert interview transcription service Stream by Mosaic—have helped expand that dataset. Users can search for things like layoff events or patents, and AlphaSense’s AI models will spit out the relevant documents.

Kokko ascribes the recent growth at his company, in part, to generative AI. Over the past couple of years, AlphaSense has incorporated a variety of large language models, including Anthropic’s recently released Claude 3, into its own in-house AI systems, which it fine-tunes with open-source models like Llama 2 and Mistral, to power new features. Where previously AlphaSense would categorize information and surface data points, now AlphaSense can help users connect the dots—summarizing information, offering industrywide analysis on what companies are saying about things like price inflation, or suggesting specific companies or competitors in sectors a user is researching.

“We’re constantly pushing what we’re able to automate for users and make them more effective at their work, and generative AI has been one of those things that really enabled big leaps in that,” Kokko says. “That has created a lot of demand—extra demand, perhaps—and a lot of attention to the use cases that we serve.”

That, plus the company’s focus on expansion into new markets like Singapore, were major contributors to AlphaSense doubling its ARR and helped attract attention from new investors. Last calendar year, amid a drought in venture capital funding to startups in sectors outside of AI, AlphaSense raised approximately $250 million in equity funding over two separate funding rounds. One of those rounds was in September, when the company raised $150 million at a $2.5 billion valuation in a deal led by BOND Capital and joined by previous investors CapitalG (one of Alphabet’s independent investing arms) and Goldman Sachs.

Darren Cohen, the CIO of Goldman Sachs’ growth equity team, sees AlphaSense as the perfect fit for their broader investment thesis: combining the infrastructure layer of AI (large language models built by companies like OpenAI or Anthropic) with a rich dataset and a specific workflow. “AlphaSense is, by far, the poster child,” Cohen tells Fortune. “It’s one of the largest positions in our fund.”

But AlphaSense still has a bit of work to do before it goes public. For one: diversifying its leadership team. There is only one woman—its new CMO—across AlphaSense’s eight-person executive team and five-person board of directors.

AlphaSense says it is currently interviewing female board candidates, and Goldman’s Cohen told Fortune that diversifying both the leadership team and board is one “of many things” AlphaSense is focused on pre-IPO. “We actually are in conversations and showing them potential candidates, some of [whom] I work closely with, which would be amazing additions. Stay tuned,” Cohen said.

The company is also updating its financial systems to handle the precise reporting necessary in the public markets and to ensure that investors can adequately compare it to other public SaaS companies, Kokko says. He anticipates that CMO Zynczak’s experience on public and private boards, in particular, will be able to guide AlphaSense on what pre- and post-IPO investors look for.

Of course, it remains to be seen exactly when the IPO markets will be deemed “open” again. While a few notable startups like Instacart and Reddit have made their debuts in recent months, listings overall have been few and far between.

Kokko won’t share an exact timeline for AlphaSense’s public listing other than to say it will depend on when the IPO market appears truly open. And given that the company raised some $250 million in equity funding just last year, Kokko says he has the capital to wait: “We’re not in a rush.”

This story was originally featured on Fortune.com