SuperMe Raises $20M to Turn Expert Knowledge Into AI Profiles

SuperMe has closed a $20 million funding round led by Forerunner's Kirsten Green, with Greylock Partners also participating. The capital is earmarked for a straightforward but demanding goal: scale a two-sided network where users ask questions and real-world experts supply the knowledge that trains their AI profiles.

The premise is familiar to anyone who has ever cold-emailed a friend-of-a-friend for career advice. SuperMe digitizes that process. Ask a question on the platform and it identifies which experts are best suited to answer, then assembles a group chat populated by their AI profiles rather than the humans themselves. Users can also bypass the group format and query an individual expert's profile directly.

Founders Casey Winters and Ludo Antonov started the company in 2024, drawing on resumes built inside some of consumer tech's most recognizable names. Winters held executive roles at Grubhub, Pinterest, and Eventbrite, and advises companies including Airbnb and Canva. Antonov led engineering teams at Pinterest, Lyft, and Whatnot. The two connected while Winters was advising Whatnot, around the time ChatGPT's rise made the idea of an AI-powered network business feel suddenly plausible.

"There are these hidden networks in Silicon Valley of who the best people are for niche skills," Winters told TechCrunch. "Our idea was: Could we make that network of who the absolute best people are more transparent, and could we take all of the key insights that are locked inside their heads and make them more readily available?"

How the Data Pipeline Works

The mechanics matter here, because the product's credibility rests entirely on data quality and trust. SuperMe scans public material, including LinkedIn, YouTube, Substack, and personal blogs, to build the foundation of an expert's AI profile. It supplements that with connected apps such as Google Drive, Notion, and Gmail, as well as meeting notetaking tools like Granola. The company has acknowledged that some of this data will include personal information and trade secrets, and says it filters that material so it never surfaces in generated answers.

That's a meaningful claim to stand behind. Expert networks have historically traded on confidentiality and liability controls precisely because the information changing hands is often sensitive, proprietary, or legally fraught. Automating that exchange through AI profiles raises the stakes on data governance rather than lowering them. SuperMe is effectively asking both sides, experts contributing knowledge and users consuming it, to trust a filtering system they can't directly audit.

The use case TechCrunch highlighted, pitch deck feedback from AI profiles modeled on venture capitalists including Mercedes Bent, Sarah Tavel, and Josh Elman, illustrates the appeal. It also illustrates the risk. A founder relying on an AI proxy of a known investor's judgment is making a bet that the proxy's advice tracks closely enough with the real person's thinking to be worth acting on. If it doesn't, the cost isn't just a bad answer. It's a founder who walked into a real fundraising meeting with false confidence.

SuperMe has also built MCP (Model Context Protocol) integration, letting users query its expert network from within existing chatbots like Claude and ChatGPT rather than switching platforms. That's a sensible distribution decision. Expert networks have always struggled with the friction of getting users to a destination site; meeting users inside tools they already use daily lowers that barrier considerably.

The company currently offers its network for free, with 13 employees and hiring underway on two fronts: improving the AI models that handle conversation and expert-matching, and expanding the roster of experts into subject areas not yet covered. Monetization plans are still being worked out. SuperMe intends to cap the number of free questions users can ask before introducing paid tiers, and it's exploring incentive structures to encourage experts to contribute more knowledge over time.

That incentive question is the real commercial hinge. Traditional expert networks like GLG or AlphaSights pay experts directly, often handsomely, for their time on calls. SuperMe's model extracts knowledge once and then monetizes it repeatedly through AI replicas, which changes the economics substantially. If experts aren't compensated in a way that scales with usage, the supply side of this marketplace could stall even as user demand grows.

Green, speaking for Forerunner, framed the investment as a bet on timing as much as technology. "I think it's one of the reasons why this is very timely, and it's good to have something like this establishing itself when we're relatively early in this journey," she said. "Because there's going to be so many lessons that are going to be learned, and they need to be shared so that people can come along in a productive way."

The Bottom Line

A $20 million Series round at this stage of a company's life is a vote of confidence, not a verdict. SuperMe has solved an interesting product problem: making tacit, hard-to-access expertise queryable at scale. It has not yet solved the harder commercial problem of compensating the humans whose knowledge powers the system, nor has it proven that users will pay once free access caps kick in.

Investors should watch two things closely over the next several quarters. First, whether SuperMe can convert free users into paying ones without the kind of churn that typically hits consumer AI products facing a paywall. Second, whether expert contributors stick around once the novelty fades and incentive mechanics become concrete rather than aspirational. Expert networks are valuable only as long as the experts keep showing up. That's the variable this round is really betting on.