Brendan Foody

Who they are

Brendan Foody is CEO and Co-Founder of Mercor — a Georgetown dropout and Thiel Fellow who started selling donuts at a 400 percent markup in eighth grade before building a $10 billion AI training data company at 22.

Person

Brendan Foody co-founded Mercor in January 2023, dropping out of Georgetown's BBA program to do it — the company was day zero when he walked away. Before that he ran Seros (2022), a project aimed at centralizing personal computing for price-performance gains in developing markets, and before that a stealth cloud consulting business (2020–2022) in Menlo Park that arbitraged AWS promotional credits. He's also listed a stint connected to Thiel Capital, which tracks: he's a Thiel Fellow, a program that pays students to leave school and build companies. The through-line is compulsive early monetization — donuts at a markup in middle school, AWS arbitrage in high school, now matching domain experts with AI labs at scale. He speaks frequently and specifically: Conversations with Tyler (October 2025) on teaching AI and the future of knowledge work, Stanford's Entrepreneurial Thought Leaders series on agentic data, The Twenty Minute VC, Lenny's Podcast, and the Notion podcast. He's active on X (@BrendanFoody), posting on Mercor's growth and AI strategy. Possibly — the dyslexia-to-entrepreneurship narrative he discusses publicly shapes how he talks about non-traditional paths into high-leverage work.

Company

Mercor publicly pivoted from AI recruiting to AI training data in June 2025 — the reframe that preceded a fivefold valuation jump to $10 billion by October 27, 2025. The Series C, $350 million led by Felicis Ventures and closed October 10, 2025, brought total funding to $492 million across 4 rounds. As of that raise, Mercor manages over 30,000 contractors and runs over $1.5 million in daily payouts. The company has a partnership with Anthropic to build datasets for Claude 4 Opus (announced May 2025) and has been actively recruiting Olympiad medalists to improve AI reasoning capabilities. In early 2026 Mercor experienced a data breach involving 4TB of stolen data via a LiteLLM supply-chain attack and has since invested in security engineering; it also faced a lawsuit from Scale AI for alleged trade secrets theft in September 2025. The new funding is earmarked for hiring 75+ employees across product engineering, business development, and risk management in 2026.

Market

Mercor competes in the AI training data and human-expert-in-the-loop space against Scale AI, Surge AI, Invisible Technologies, Labelbox, and Turing AI — a market projected to reach $1.6 billion. The competitive dynamic is sharpening: Meta's investment in Scale AI is shifting lab partnerships, and Scale AI has already sued Mercor for trade secrets theft, signaling that rivalry is no longer polite. Regulatory exposure is growing too — the EU AI Act, NYC's AEDT rules, and data privacy frameworks all touch Mercor's model, and the 2026 breach may accelerate classification of AI middleware as critical infrastructure.

Network

No direct edges are available in the claims. Brendan's known public interlocutors include Tyler Cowen (Conversations with Tyler, October 2025), Harry Stebbings (The Twenty Minute VC, 2025), and Lenny Rachitsky (Lenny's Podcast, 2025). His co-founders at Mercor are Adarsh Hiremath (CTO) and Surya Midha (COO), all three of whom became the youngest self-made billionaires in 2025 at age 22.

Adarsh Hiremath· CTO and Co-Founder, Mercor
Surya Midha· COO and Co-Founder, Mercor
Tyler Cowen· Host, Conversations with Tyler; Economist, George Mason University
Harry Stebbings· Host, The Twenty Minute VC
Lenny Rachitsky· Host, Lenny's Podcast

How they likely show up

Serial founding from age 13 (donuts), through high school (AWS arbitrage), to Georgetown (Seros), to dropout (Mercor) → operates on a continuous build loop; waiting and watching is not a natural mode.
Thiel Fellowship and Georgetown dropout → has explicitly chosen speed and ownership over credentials; likely skeptical of slow institutional processes.
Active across Conversations with Tyler, Stanford STVP, The Twenty Minute VC, Lenny's Podcast, and Notion podcast in 2025 alone → comfortable in high-signal public forums, not just friendly press; engages with intellectually rigorous interlocutors.
Pivoted Mercor's public identity from recruiting to AI training data mid-2025, ahead of the valuation inflection → willing to reframe the narrative sharply when the underlying reality has shifted; not attached to original positioning.
Possibly — dyslexia narrative he discusses publicly suggests he is drawn to pattern-recognition and first-principles thinking over linear, text-heavy analysis.
Role type is founder-CEO from day one with no prior large-company employment → decisions come from owner's conviction, not committee; likely moves fast and expects others to keep up.

Conversation tips

→ Reference the June 2025 pivot specifically — from recruiting to AI training data — and ask what that reframe unlocked internally; it's the strategic move that changed the company's trajectory and he'll have a real answer.
→ He's appeared with Tyler Cowen and Lenny Rachitsky, both of whom go deep and specific. Match that register — come with a precise question, not a broad one.
→ The Scale AI lawsuit and the 2026 data breach are live issues; don't pretend they don't exist, but don't make them the opener — let him bring them in on his terms.
→ He built a donut business at a 400 percent markup in eighth grade: this isn't color, it's the founding logic of Mercor. Connecting early arbitrage instinct to how he prices and structures Mercor's marketplace will resonate.

Toolbox

Openers

Open on the June 2025 pivot — Mercor publicly reframed from AI recruiting to AI training data company just months before quintupling its valuation to $10 billion. That sequencing is deliberate; asking what drove the timing of the public reframe will get a real strategic answer.
Reference his Conversations with Tyler episode (October 2025) on teaching AI and the future of knowledge work — Tyler's audience is economists and intellectuals, and Foody chose to go deep on labor market theory there, not just founder storytelling. Engaging on those ideas signals you did the work.
Bring up the Anthropic partnership for Claude 4 Opus datasets (May 2025) — it's the clearest public proof point of Mercor's shift from marketplace to model-training infrastructure, and it anchors a conversation about where the real value in the AI supply chain sits.

Discovery questions

When you pivoted Mercor's identity from recruiting to AI training data in mid-2025, what internally had already changed — was the pivot catching up to the revenue mix, or did the reframe drive the product direction?
You've talked publicly about matching domain experts with AI labs for evals and training data — how do you think about the durability of that human-in-the-loop layer as models get more capable?
The Scale AI lawsuit and the LiteLLM breach both landed within months of each other in late 2025 and early 2026 — how has that compressed timeline changed how you think about security and legal infrastructure at Mercor's current scale?

Avoid

Don't treat Mercor as primarily a recruiting or HR-tech company — Foody publicly reframed the company as an AI training data business in mid-2025, and leading with the old framing signals you haven't followed the story.

Make it yours