Pinegap, an AI-powered equity research platform designed for institutional buy-side investors that its founder says is “like having a junior analyst,” has closed an $8 million Series A round led by Stellaris Venture Partners.
The company, which charges $6,000 per user on a subscription basis, builds proactive custom AI agents tuned to a fund’s investment style, private data, and output formats. Rather than waiting for an analyst to ask a question, the agents work in the background, automatically sending earnings previews, company primers, and thesis updates to an analyst's inbox on a set schedule or whenever something happens in the market.
The product launched commercially in the first quarter of 2025. The company says it has deployed over 1,000 agents across more than 100 institutional clients, generating upwards of 50,000 research reports per month. Annual recurring revenue has since doubled, and co-founder Ankit Varmani says he expects revenue to triple by the end of the year.
Pinegap's clients include "five of the top 10 multi-managers in New York, and some of the most recognizable names in the mutual fund industry," Varmani said, though client agreements largely prevent him from naming names. The client base is roughly split between long/short and long-only funds, with the company also expanding into boutique RIA firms.
The company has 21 employees mostly in technical roles, and plans to use the new funding to expand customer support, marketing, sales, and engineering. It has already hired six former analysts to support its sales process.
Existing backers Inventus, Silicon Valley Quad, and DeVC also participated in the financing. New York-based Pinegap’s valuation is "more than triple" its $2.5 million seed round around the time it was founded in February 2024, though Varmani declined to disclose exact figures.
From Wall Street to Founder
Varmani spent nearly 15 years as an equity analyst before starting the company, beginning his career in industrials research on the institutional sell side at J.P. Morgan. He left in 2014 for the buy side, spending eight years as a long/short equity analyst at HHR Asset Management until the end of 2022.
Covering technology stocks such as Nvidia at HHR is where he first got interested in AI.
“You could see a lot of parallels between what I do as an analyst and what this technology can offer,” he told This Week in Fintech in an interview.
The founding idea traces back to an assignment he gave his eventual co-founder, Deepak Sharma, an IIT alumnus who was then preparing to learn the craft of equity research. While working through those tasks, Sharma began turning to early large language models to solve them. The two quickly realized that the most repetitive, time-consuming parts of equity research could be automated with AI and built Pinegap to give every buy-side fund these tools.
“It was my plan to basically think about AI-focused equity research, and what Deepak brought to the table was the technical know-how,” said Varmani, who then left his job to work on the idea full time, building the product through 2023 and formally registering Pinegap, with Sharma as CEO, in January 2024.
Left to right: Pinegap co-founders Ankit Varmani and Deepak Sharma/Credit: Pinegap
What Pinegap Does
The platform is built around core analyst workflows such as sourcing new ideas, ramping up on new companies, preparing for and reacting to earnings, tracking conferences, and managing news flow – with a dedicated AI module for each. It draws on earnings calls and SEC filings as well as newer additions like podcast, Twitter, and Reddit data. The company has also introduced an “MCP offering” that functions “like a modern-day API,” letting larger funds' internal tech teams build on top of Pinegap's data rather than “reinventing the wheel in-house,” according to Varmani.
"The whole idea was how do I make myself more efficient as an analyst using whatever latest and greatest AI has to offer," he said. "This tool is purpose-built for buy-side equity research analysts... This is entirely what we do."
Varmani argues that general-purpose AI tools fall short for institutional use because they pull from “everything out there on the internet,” raising the risk of inaccurate answers. Pinegap's integration, by contrast, connects directly to a fund's own data.
“You never have a problem of hallucination and stuff like that on the platform,” Varmani said. He pointed to the platform's proactive alerts as a feature that tends to win analysts over. If a company on their watch list reports earnings, for example, Pinegap automatically emails them the "key debates" from the call without being asked.
"It's truly like having a junior analyst," he said.
Varmani sees two types of competitors: legacy incumbents that have been “slow to catch up on AI,” and newer AI startups that have largely built chatbot-style tools.
“Not everything can be addressed using a chatbot,” he said. “Sometimes you just don't know what you don't know.”
Pinegap's differentiators, he said, are being purpose-built for a single niche, its growing base of proprietary data, and its flexibility in serving both "buy" and "build" customers.
Scale, Team, and Clients
“We work very closely with our customers, even in the very initial stages of the sales cycle, to truly understand what they are doing,” Varmani said. Although Pinegap is headquartered in New York, it also has a technology team in Bangalore and serves clients in London, Singapore, and Hong Kong in addition to the U.S.
Alok Goyal, a partner at Stellaris Venture Partners, said the deal fits the firm's thesis on vertical AI – backing "domain-specific, workflow-integrated platforms that solve nuanced, non-commoditized industry problems."
Goyal also pointed to Pinegap's proactive design as a key edge over “pull-based” tools where “the analyst asks, the tool answers.”
Pinegap instead “ingests a fund's actual context” and proactively delivers earnings previews, news summaries, and thesis tracking the way each fund builds them, he said, letting analysts “focus on making the actual judgment rather than the grunt work.”
He added that the real barrier to competition isn't the AI itself.
“The hard part isn't the intelligence layer,” Goyal told This Week in Fintech. “It's the workflow layer, and that's domain knowledge earned one customer at a time, which is very hard to replicate.”
