LoginSign Up
Sign Up

New to SuperMe?

Sign up free to work on your questions with people who've been there, like Ludo.

Sign up for freeLog in

Ludo Antonov

Co-Founder and Chief Token Officer, SuperMe

522K insights
Visible to Ludo
Ludo won't be able to see or reply to your conversation in incognito mode

About

I'm Ludo Antonov, Co-Founder of SuperMe. We're building the most useful network on the planet - (it's a human dash, relax) one that makes real human expertise accessible in ways LLMs alone can't. Before SuperMe, I was VP/Head of Engineering at Whatnot, leading the engineering team through its breakout phase as one of the fastest-growing consumer marketplaces. At Lyft, I led 350+ engineers across rider, driver, marketplace, and growth. At Pinterest, I built the Growth Engineering team from inception to 100+ engineers, helping take the platform from 20M to 200M+ users. The common thread is that I'm drawn to inflection points - the moment a product goes from "we think this might work" to knowing it's working. I've spent my career building consumer products at scale, and growth has always been a natural extension of getting the product right. That same instinct is behind SuperMe. Most expertise lives in people's heads, in pattern recognition earned by doing the work. We're making that legible and useful.

San Francisco, California

Work Experience

Co-Founder at
SuperMe

May 2024 - Present (2y 6m)

SuperMe is on a mission to index collective human knowledge and build the most useful network on the planet.

VP, Head of Engineering at

Dec 2021 - May 2024 (2y 5m)

Led the engineering team at Whatnot. Building the future of commerce.

Head of Growth & Product Engineering at

May 2019 - Dec 2021 (2y 7m)

Leading the engineering organization 350+ focused on building Lyft's core product across driver, growth, marketplace, and rider.

Head of Growth Engineering at

Feb 2013 - May 2019 (6y 3m)

Built the Pinterest Growth Engineering team from inception to 100+ engineers in the process of taking Pinterest from 20mln to 200+mln users worldwide. Bootstrapped many of the areas of growth - seo, email/notifs, social, performance, user activation, user acquisition. Started Open Source initiative at Pinterest.

Timeline

Ludo Antonov

Josh Elman

@joshelman

Love that. Step-by-step. Give-then-ask.

View on 𝕏
Ludo Antonov

Josh Elman

@joshelman

Wait what?? Double digit wins? Can you share what the advice is helping with so I can make my advice even better?

View on 𝕏
Ludo Antonov
View on LinkedIn
Ludo Antonov
4d

Gul Saeed Khattak

@Ldataengineer

What does it mean to ask an expert a question when the expert isn’t there? SuperMe raised $20M (Forerunner leading, Greylock participating) for AI profiles of real experts. The mechanics: it ingests an expert’s public output (LinkedIn, YouTube, Substack, blogs) plus, if they connect them, Drive, Notion, Gmail and Granola notes. Then a router picks which experts’ profiles should answer your question and drops them into one group chat. I think the interesting part isn’t the AI clones. Delphi has done expert “digital minds” for a while. It’s the routing. The product is basically a search engine where the documents are people’s reasoning, and the ranking problem is “whose perspective is relevant here.” That makes it closer to retrieval than to a chatbot. And retrieval has a known failure mode: confident synthesis over thin evidence. One early tester found a throwaway remark presented as something he “often asks.” So what I’d want to see is attribution at the sentence level. For each claim an expert’s AI makes, show the source it came from, and say clearly when the expert never took a position. Reports say it already flags topics where an expert hasn’t taken a position, which is the right instinct. Free for now, with a question cap and payment later. No current user numbers are public. https://techcrunch.com/2026/10/29/superme-raises-20m-for-its-network-of-ai-expert-profiles/

What does it mean to ask an expert a question when the expert isn’t there?

SuperMe raised $20M (Forerunner leading, Greylock participating) for AI profiles of real experts.

The mechanics: it ingests an expert’s public output (LinkedIn, YouTube, Substack, blogs) plus, if they connect them, Drive, Notion, Gmail and Granola notes. Then a router picks which experts’ profiles should answer your question and drops them into one group chat.

I think the interesting part isn’t the AI clones. Delphi has done expert “digital minds” for a while. 
It’s the routing. 
The product is basically a search engine where the documents are people’s reasoning, and the ranking problem is “whose perspective is relevant here.”

That makes it closer to retrieval than to a chatbot. 

And retrieval has a known failure mode: confident synthesis over thin evidence. One early tester found a throwaway remark presented as something he “often asks.”

So what I’d want to see is attribution at the sentence level. For each claim an expert’s AI makes, show the source it came from, and say clearly when the expert never took a position. 
Reports say it already flags topics where an expert hasn’t taken a position, which is the right instinct.

Free for now, with a question cap and payment later. No current user numbers are public.
https://techcrunch.com/2026/10/29/superme-raises-20m-for-its-network-of-ai-expert-profiles/
View on 𝕏
🔥
1
Ludo Antonov

Unknown

@twitter.com

View on 𝕏